Hugging Face's logo

Datasets:
PolyU-ChenLab
/
ETBench

Modalities:
Text
Formats:
text
Size:
< 1K
ArXiv:
Libraries:
Datasets
License:
Dataset card Viewer Files Community
1
yeliudev commited on
Commit
6b67d78
•
1 Parent(s): fef5c15

Add dataset files

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. annotations/etbench_txt_v1.0.json +0 -0
  2. annotations/etbench_vid_v1.0.json +0 -0
  3. annotations/txt/dvc_hirest.json +1 -0
  4. annotations/txt/dvc_youcook2.json +0 -0
  5. annotations/txt/eca_charades_sta.json +0 -0
  6. annotations/txt/epm_eg4d_nlq.json +0 -0
  7. annotations/txt/evs_summe.json +1 -0
  8. annotations/txt/evs_tvsum.json +1 -0
  9. annotations/txt/gvq_qa_ego4d.json +0 -0
  10. annotations/txt/rar_perception_test.json +0 -0
  11. annotations/txt/rvq_star.json +0 -0
  12. annotations/txt/slc_cross_task.json +0 -0
  13. annotations/txt/slc_ht_step.json +0 -0
  14. annotations/txt/tal_perception_test.json +0 -0
  15. annotations/txt/tal_thumos14.json +1 -0
  16. annotations/txt/tal_thumos15.json +0 -0
  17. annotations/txt/tem_perception_test.json +0 -0
  18. annotations/txt/tem_qvhighlights.json +0 -0
  19. annotations/txt/tvg_charades_sta.json +0 -0
  20. annotations/txt/tvg_qvhighlights.json +0 -0
  21. annotations/txt/vhd_qvhighlights.json +0 -0
  22. annotations/txt/vhd_youtube_highlights.json +0 -0
  23. annotations/vid/dvc_hirest.json +1 -0
  24. annotations/vid/dvc_youcook2.json +0 -0
  25. annotations/vid/eca_charades_sta.json +0 -0
  26. annotations/vid/epm_eg4d_nlq.json +0 -0
  27. annotations/vid/evs_summe.json +1 -0
  28. annotations/vid/evs_tvsum.json +1 -0
  29. annotations/vid/gvq_qa_ego4d.json +0 -0
  30. annotations/vid/rar_perception_test.json +0 -0
  31. annotations/vid/rvq_star.json +0 -0
  32. annotations/vid/slc_cross_task.json +0 -0
  33. annotations/vid/slc_ht_step.json +0 -0
  34. annotations/vid/tal_perception_test.json +0 -0
  35. annotations/vid/tal_thumos14.json +1 -0
  36. annotations/vid/tal_thumos15.json +0 -0
  37. annotations/vid/tem_perception_test.json +0 -0
  38. annotations/vid/tem_qvhighlights.json +0 -0
  39. annotations/vid/tvg_charades_sta.json +0 -0
  40. annotations/vid/tvg_qvhighlights.json +0 -0
  41. annotations/vid/vhd_qvhighlights.json +0 -0
  42. annotations/vid/vhd_youtube_highlights.json +0 -0
  43. compress_videos.py +51 -0
  44. evaluation/compute_metrics.py +971 -0
  45. evaluation/requirements.txt +5 -0
  46. evaluation/subset.json +1 -0
  47. videos/charades_sta.tar.gz +3 -0
  48. videos/cross_task.tar.gz +3 -0
  49. videos/ego4d_nlq.tar.gz +3 -0
  50. videos/hirest.tar.gz +3 -0
annotations/etbench_txt_v1.0.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/etbench_vid_v1.0.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/txt/dvc_hirest.json ADDED
@@ -0,0 +1 @@
 
 
1
+ [{"version": 1.0, "idx": 0, "task": "dvc", "source": "hirest", "video": "hirest/5V3dI2zp1xA_50_121.mp4", "duration": 71.0, "tgt": [[0, 8], [8, 17], [17, 30], [30, 54], [54, 71]], "g": ["take grinned oats flour", "add carrot and chopped spinach", "add coriander and green chilly", "add salt and water mix well", "heat pan and add fruit salt"], "q": "You are given a video about 'make oatmeal pancake mix'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 1, "task": "dvc", "source": "hirest", "video": "hirest/TcB42a05yzg_33_156.mp4", "duration": 123.0, "tgt": [[0, 11], [11, 18], [18, 19], [19, 29], [29, 37], [37, 51], [51, 100], [100, 123]], "g": ["take thee plate with three candlesticks", "color all with color spray", "add glue to candle sstick", "paste the plate on candle stick", "repeat until the three layer stand completed", "place cup cakes on each plate", "change the candlestick size to our need", "decorate the top cupcake with toy"], "q": "You are given a video about 'make a cupcake stand'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 2, "task": "dvc", "source": "hirest", "video": "hirest/2YZHJOLuA8o_63_216.mp4", "duration": 153.0, "tgt": [[0, 11], [11, 41], [41, 60], [60, 73], [73, 85], [85, 96], [96, 107], [107, 113], [113, 133], [133, 153]], "g": ["take seasonal pan", "spray it and put leafs onit", "spread leafs on pan", "cover empty places on leaf", "add spray on top", "add little bit salt on salt", "add pepper on leaf", "put pan on oven", "cook it for six minutes", "transfer cooked leafs on bowl"], "q": "You are given a video about 'make easy and healthy chips'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 3, "task": "dvc", "source": "hirest", "video": "hirest/Gb-uDDzFRLE_17_49.mp4", "duration": 32.0, "tgt": [[0, 15], [15, 23], [23, 32]], "g": ["take bottom and measure size", "cut straight and well", "cut as slide way"], "q": "You are given a video about 'make giant foam candy canes'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 4, "task": "dvc", "source": "hirest", "video": "hirest/Sb9XLGPJ1yk_25_92.mp4", "duration": 67.0, "tgt": [[0, 11], [11, 21], [21, 37], [37, 43], [43, 51], [51, 56], [56, 63], [63, 67]], "g": ["take dishwashing soap", "take olive oil", "take plate", "pour soap on plate", "pour olive oil on plate", "mix with brush", "lather on hand", "squeeze out makeup"], "q": "You are given a video about 'make your own makeup brush cleanser'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 5, "task": "dvc", "source": "hirest", "video": "hirest/FcRrnEpE4x8_165_439.mp4", "duration": 274.0, "tgt": [[0, 3], [3, 40], [40, 72], [72, 92], [92, 136], [136, 163], [163, 173], [173, 186], [186, 217], [217, 267], [267, 274]], "g": ["zero out the container", "put ink", "mix it around", "spray on the palette", "pull on ink", "flash the garment", "do over print", "do quick flash", "put over print", "cure it", "do a stretch test"], "q": "You are given a video about 'do screen printing on fabric'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 6, "task": "dvc", "source": "hirest", "video": "hirest/-DiQNHSyUfs_34_108.mp4", "duration": 74.0, "tgt": [[0, 11], [11, 18], [18, 38], [38, 58], [58, 74]], "g": ["take some water in plastic bag", "mix vinegar with water", "shut the bag but corner open", "take baking soda in tissue paper", "slip tissue inbag and seal it"], "q": "You are given a video about 'make a plastic bag explode'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 7, "task": "dvc", "source": "hirest", "video": "hirest/gqEbqMKoh9I_33_104.mp4", "duration": 71.0, "tgt": [[0, 18], [18, 37], [37, 56], [56, 67], [67, 71]], "g": ["take paper and put soda onit", "fold paper with soda", "put vinegar with sandwich bag", "put vinegar paper on bag", "shack it well"], "q": "You are given a video about 'make a plastic bag explode'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 8, "task": "dvc", "source": "hirest", "video": "hirest/S6N5RpvChWs_32_113.mp4", "duration": 81.0, "tgt": [[0, 24], [24, 62], [62, 71], [71, 72], [72, 81]], "g": ["take water with vinegar in plastic bags", "take baking soda in tissue paper", "fold paper well", "put it into the bag", "keep away to explode"], "q": "You are given a video about 'make a plastic bag explode'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 9, "task": "dvc", "source": "hirest", "video": "hirest/hHgiALvGm5Y_34_158.mp4", "duration": 124.0, "tgt": [[0, 18], [18, 30], [30, 49], [49, 60], [60, 85], [85, 108], [108, 124]], "g": ["take wanted materials", "put baking soda on paper", "fold soda on bundle", "put soda packet on liquid", "put pocket on ground", "it out on bubbles", "repeat same process on another bag"], "q": "You are given a video about 'make a plastic bag explode'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 10, "task": "dvc", "source": "hirest", "video": "hirest/LKSvFmXLQA0_27_46.mp4", "duration": 19.0, "tgt": [[0, 10], [10, 19]], "g": ["make the fire to use cook", "place large logs into bottom"], "q": "You are given a video about 'use a fire pit'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 11, "task": "dvc", "source": "hirest", "video": "hirest/5OVxioAu6IY_46_163.mp4", "duration": 117.0, "tgt": [[0, 19], [19, 44], [44, 117]], "g": ["arrange the hardware parts", "take the air pipe", "take the nuts and bolds"], "q": "You are given a video about 'use a fire pit'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 12, "task": "dvc", "source": "hirest", "video": "hirest/aHDGMOQgHUE_35_99.mp4", "duration": 64.0, "tgt": [[0, 28], [28, 64]], "g": ["make a fire with coal", "check the gas"], "q": "You are given a video about 'use a fire pit'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 13, "task": "dvc", "source": "hirest", "video": "hirest/Tu9foJWCkgs_47_82.mp4", "duration": 35.0, "tgt": [[0, 10], [10, 16], [16, 25], [25, 35]], "g": ["remove lid", "put handle", "pick grill to put stones", "put the lid back"], "q": "You are given a video about 'use a fire pit'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 14, "task": "dvc", "source": "hirest", "video": "hirest/PtbGXfb6B1I_46_209.mp4", "duration": 163.0, "tgt": [[0, 115], [115, 126], [126, 140], [140, 163]], "g": ["dig the soil", "put the stick inside big hole", "adjust the stick by small hole", "fire the stick"], "q": "You are given a video about 'use a fire pit'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 15, "task": "dvc", "source": "hirest", "video": "hirest/wfiAtfED4K8_41_159.mp4", "duration": 118.0, "tgt": [[0, 8], [8, 13], [13, 24], [24, 30], [30, 39], [39, 59], [59, 118]], "g": ["put the logs into the firepit", "spread the larger logs", "put logs crisscross on top", "put the next size", "put tinder bundle on them", "add some smaller brunches", "start the fire"], "q": "You are given a video about 'use a fire pit'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 16, "task": "dvc", "source": "hirest", "video": "hirest/4zQJzWvRJRA_93_285.mp4", "duration": 192.0, "tgt": [[0, 4], [4, 8], [8, 192]], "g": ["put fence post", "add concrete footing", "keep it above soil"], "q": "You are given a video about 'install a privacy fence'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 17, "task": "dvc", "source": "hirest", "video": "hirest/mqJs3zrKACU_75_246.mp4", "duration": 171.0, "tgt": [[0, 7], [7, 14], [14, 17], [17, 18], [18, 22], [22, 32], [32, 38], [38, 40], [40, 43], [43, 49], [49, 57], [57, 59], [59, 68], [68, 72], [72, 79], [79, 91], [91, 93], [93, 109], [109, 125], [125, 143], [143, 159], [159, 171]], "g": ["heat oil in pan", "add mustard seeds", "add garlic and ginger", "add green chili", "add curry leaves", "saute them", "add turmeric powder", "add beets", "add salt", "saute them again", "cover it to cook", "put fenugreek seeds in pan", "roast it", "add mustard seeds", "roast it again", "crush them", "add crushed seeds into beets", "add asafetida", "cook it", "add red chili powder", "add white vinegar", "stir it"], "q": "You are given a video about 'make spicy sauteed beets with mustard and lemon'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 18, "task": "dvc", "source": "hirest", "video": "hirest/HjEf_JHzqe8_40_157.mp4", "duration": 117.0, "tgt": [[0, 6], [6, 8], [8, 17], [17, 35], [35, 41], [41, 48], [48, 57], [57, 63], [63, 71], [71, 76], [76, 94], [94, 108], [108, 117]], "g": ["remove the front grill", "check condenser", "unbolt the top", "unbolt the lines", "put paper tower in lines", "remove the condenser", "put pag oil", "put new condenser", "put ac oil", "put bolt back", "tighten it up", "remove air out of system", "fill in the refrigerant"], "q": "You are given a video about 'fix your car's air conditioner'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 19, "task": "dvc", "source": "hirest", "video": "hirest/6nnKM_y_vn8_81_358.mp4", "duration": 277.0, "tgt": [[0, 7], [7, 18], [18, 63], [63, 76], [76, 99], [99, 112], [112, 127], [127, 147], [147, 183], [183, 201], [201, 215], [215, 229], [229, 263], [263, 273], [273, 277]], "g": ["add browned beef", "put cooked green bell pepper", "add cooked onions", "add chopped cilantro", "add garlic powder", "add cumin", "add chili powder", "add hot sauce", "mix it around with beans", "add diced tomatoes", "add tomato sauce", "add sugar", "mix it all", "put instant pot on saute", "put it on keep warm later"], "q": "You are given a video about 'make glazed beans'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 20, "task": "dvc", "source": "hirest", "video": "hirest/38H8OKlNA9g_55_178.mp4", "duration": 123.0, "tgt": [[0, 30], [30, 31], [31, 41], [41, 53], [53, 60], [60, 89], [89, 95], [95, 123]], "g": ["open the box of smoker", "check the filling status of charcoal", "add more wood chips if required", "close the box after filling", "ensure air can get around the wood", "check the fire of smoker", "cook the meat well", "check the temperature between the correct degree"], "q": "You are given a video about 'use a smoker'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 21, "task": "dvc", "source": "hirest", "video": "hirest/8KWG-HxHcbE_126_412.mp4", "duration": 286.0, "tgt": [[0, 13], [13, 18], [18, 29], [29, 31], [31, 47], [47, 56], [56, 67], [67, 80], [80, 98], [98, 126], [126, 130], [130, 141], [141, 159], [159, 173], [173, 182], [182, 208], [208, 226], [226, 244], [244, 273], [273, 286]], "g": ["preheat oven", "add vinegar at papaya slices", "spread at gently", "add vinegar for tomato slices", "put it on tray", "put tray on oven", "add oil on pan", "add vegetable on pan", "add salt and mix well", "cook squash until fork tender", "add mixer on jar", "add chicken broken on mixer", "put mixer on pot add chickenoil", "add chock salt and pepper", "add chocki on pot", "add chicken drop and banana pieces", "add chopped cease", "add herbs on pot", "mix all well", "add doll and mix well"], "q": "You are given a video about 'make chorizo'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 22, "task": "dvc", "source": "hirest", "video": "hirest/5Jn7OuXb0FA_25_83.mp4", "duration": 58.0, "tgt": [[0, 2], [2, 10], [10, 23], [23, 35], [35, 51], [51, 58]], "g": ["take your felt", "draw your stencil", "cut it out", "lay it out first", "glue the gems", "glue the pin"], "q": "You are given a video about 'make a brooch'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 23, "task": "dvc", "source": "hirest", "video": "hirest/NULTB33w3Bw_56_282.mp4", "duration": 226.0, "tgt": [[0, 5], [5, 11], [11, 24], [24, 71], [71, 95], [95, 107], [107, 114], [114, 144], [144, 161], [161, 193], [193, 226]], "g": ["put acid on glass", "mix until dissolved set aside", "add glue into mixing bowl:", "add foaming soap into bowl", "add shaving cream into bowl", "mix everything together", "add corn starch into mixture", "add pumps of lotion", "add food coloring into bowl", "add borax solution into bowl", "mix and get right consistency"], "q": "You are given a video about 'make fluffy slime'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 24, "task": "dvc", "source": "hirest", "video": "hirest/xcrmJR_RVIw_31_156.mp4", "duration": 125.0, "tgt": [[0, 79], [79, 125]], "g": ["arrange the flowers in the clutch bags", "add a small yellow petals"], "q": "You are given a video about 'dry flowers using a bag'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 25, "task": "dvc", "source": "hirest", "video": "hirest/NPcyRoWVfL4_43_418.mp4", "duration": 375.0, "tgt": [[0, 25], [25, 59], [59, 91], [91, 150], [150, 185], [185, 192], [192, 237], [237, 266], [266, 375]], "g": ["soak fibers in water", "add white vinegar", "add food colors", "mix it", "put in microwave", "take it out", "let it cool", "take out fibers", "wash them"], "q": "You are given a video about 'set vegetable dyes in clothing'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 26, "task": "dvc", "source": "hirest", "video": "hirest/qnN_OLNAHvc_173_230.mp4", "duration": 57.0, "tgt": [[0, 8], [8, 19], [19, 23], [23, 36], [36, 50], [50, 57]], "g": ["assemble the jig", "align it", "set the blade", "check the depth", "spray water on blade while cutting", "keep cords dry"], "q": "You are given a video about 'cut concrete'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 27, "task": "dvc", "source": "hirest", "video": "hirest/qY9CerJY7FU_80_322.mp4", "duration": 242.0, "tgt": [[0, 37], [37, 77], [77, 179], [179, 242], [242, 242]], "g": ["put egg in the bowl", "apply over the face", "put tissue over it", "dry it", "remove the tissue"], "q": "You are given a video about 'make egg strips'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 28, "task": "dvc", "source": "hirest", "video": "hirest/lpyvvneQImU_81_236.mp4", "duration": 155.0, "tgt": [[0, 2], [2, 8], [8, 14], [14, 21], [21, 37], [37, 43], [43, 44], [44, 46], [46, 51], [51, 56], [56, 60], [60, 70], [70, 96], [96, 155]], "g": ["pre-heat a pan", "add oil to the pan", "add white onion", "sautee the onion", "add garlic", "add apple cider vinegar", "add salt", "add pepper", "add butter", "bring mixture to a little boil", "add brown sugar", "add yellow mustard", "reduce heat and simmer until thickened", "drain mustard sauce"], "q": "You are given a video about 'make a simple mustard sauce'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 29, "task": "dvc", "source": "hirest", "video": "hirest/Ti9ibiKSpf8_13_79.mp4", "duration": 66.0, "tgt": [[0, 11], [11, 25], [25, 33], [33, 39], [39, 66]], "g": ["add mustard", "add honey", "add garlic", "add sriracha sauce", "mix it"], "q": "You are given a video about 'make a simple mustard sauce'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 30, "task": "dvc", "source": "hirest", "video": "hirest/cQ0namNxwWA_99_152.mp4", "duration": 53.0, "tgt": [[0, 9], [9, 12], [12, 13], [13, 16], [16, 22], [22, 26], [26, 30], [30, 34], [34, 35], [35, 53]], "g": ["prepare the ingredients", "beat the egg with cinnamon", "add nutmeg with allspice to taste", "spread some almond meal on plate", "place ham on pineapple piece", "dip that to the beaten egg", "coat pineapple with almond meal", "coat ham with almond meal", "place on heated fry pan with oil", "cook until it has browned"], "q": "You are given a video about 'make pineapple fritters'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 31, "task": "dvc", "source": "hirest", "video": "hirest/P2b7fEpIBIg_48_245.mp4", "duration": 197.0, "tgt": [[0, 50], [50, 97], [97, 125], [125, 149], [149, 180], [180, 197]], "g": ["mix the ingredients", "beat them well", "mix dry and wet ingredients", "put them in the try", "decorate with nuts", "put them in oven"], "q": "You are given a video about 'make pineapple fritters'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 32, "task": "dvc", "source": "hirest", "video": "hirest/WcyDbvsftkI_58_261.mp4", "duration": 203.0, "tgt": [[0, 8], [8, 180], [180, 203]], "g": ["cut all pieces at once", "get a pattern", "paste all the pieces through the pattern"], "q": "You are given a video about 'foundation piece a quilt block'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 33, "task": "dvc", "source": "hirest", "video": "hirest/o69yX5dzvDI_70_301.mp4", "duration": 231.0, "tgt": [[0, 65], [65, 134], [134, 231]], "g": ["fix the design paper", "separate the ribbons", "arrange it like a paper design"], "q": "You are given a video about 'foundation piece a quilt block'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 34, "task": "dvc", "source": "hirest", "video": "hirest/2nkCdciXDa8_55_423.mp4", "duration": 368.0, "tgt": [[0, 79], [79, 91], [91, 109], [109, 120], [120, 137], [137, 162], [162, 178], [178, 194], [194, 197], [197, 200], [200, 213], [213, 216], [216, 255], [255, 309], [309, 346], [346, 353], [353, 368]], "g": ["cut the fabric", "put on template", "put paper clips", "make back stitches", "remove paper clips", "whip stitch them together", "do another one", "attach background piece", "remove the template", "fold it in half", "do the seams", "merge them", "sew the line", "attach pebble piece", "sew the edges", "fill the gaps", "make hexagon edges"], "q": "You are given a video about 'foundation piece a quilt block'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 35, "task": "dvc", "source": "hirest", "video": "hirest/5JpT5RcuFUk_115_251.mp4", "duration": 136.0, "tgt": [[0, 2], [2, 8], [8, 20], [20, 24], [24, 28], [28, 34], [34, 59], [59, 66], [66, 79], [79, 86], [86, 136]], "g": ["get plastic glass table", "put light underneath", "flip over the copy", "use a ruler", "use sharpie to draw", "continue to draw lines", "transfer the numbers of patches", "use double mirror", "lay it on inside", "piece the foundation", "sew the foundation sections"], "q": "You are given a video about 'foundation piece a quilt block'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 36, "task": "dvc", "source": "hirest", "video": "hirest/9Bc1mU0LMF8_82_385.mp4", "duration": 303.0, "tgt": [[0, 12], [12, 37], [37, 100], [100, 106], [106, 129], [129, 145], [145, 166], [166, 173], [173, 177], [177, 179], [179, 189], [189, 202], [202, 231], [231, 238], [238, 271], [271, 274], [274, 301], [301, 303]], "g": ["put scrap", "flip it over", "stitch along line", "trim it", "fold it and iron", "attach piece along seam line", "cut extra fabric", "hold over piece", "add the other piece", "trim the seam", "press with iron", "add other pieces", "attach wide black borders to sides", "attach borders on top and bottom", "draw line to stitch", "flip it over", "tear off paper", "press it"], "q": "You are given a video about 'foundation piece a quilt block'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 37, "task": "dvc", "source": "hirest", "video": "hirest/LX_VHvjunFM_4_57.mp4", "duration": 53.0, "tgt": [[0, 17], [17, 51], [51, 53]], "g": ["things to make krispie buns", "put water in pan,mix chocloate", "put the mix in cups"], "q": "You are given a video about 'make rice krispie buns'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 38, "task": "dvc", "source": "hirest", "video": "hirest/opGK2SjlWJU_57_103.mp4", "duration": 46.0, "tgt": [[0, 25], [25, 40], [40, 46]], "g": ["cut the edges of turnip", "cut into small pieces", "cook with vinegar pepper salt"], "q": "You are given a video about 'cook turnips'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 39, "task": "dvc", "source": "hirest", "video": "hirest/9pybtCnEqEQ_28_125.mp4", "duration": 97.0, "tgt": [[0, 3], [3, 26], [26, 65], [65, 82], [82, 92], [92, 97], [97, 97]], "g": ["make grass using bottom of the", "using pencil draw grass", "paint the grass and dry", "paint the flowers with different colors", "put together flowers and grass", "draw leaves", "leave some message in the card"], "q": "You are given a video about 'make a vintage gift tag'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 40, "task": "dvc", "source": "hirest", "video": "hirest/VvpdSid8ruo_74_124.mp4", "duration": 50.0, "tgt": [[0, 7], [7, 17], [17, 27], [27, 30], [30, 34], [34, 50]], "g": ["paste wrapping paper backside to sheet", "cut the outlined deign", "paste the brown paper in middle", "punch the top of tag", "tie lace in the hole", "write message in the brown sheet"], "q": "You are given a video about 'make a vintage gift tag'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 41, "task": "dvc", "source": "hirest", "video": "hirest/UK3-ZpnogDc_103_347.mp4", "duration": 244.0, "tgt": [[0, 9], [9, 24], [24, 28], [28, 37], [37, 61], [61, 74], [74, 165], [165, 193], [193, 216], [216, 244]], "g": ["put paper", "put on craft bond", "stick them", "mark it", "put glitter tape", "make cut outs", "repeat the process", "punch holes", "take threads", "tie threads on tags"], "q": "You are given a video about 'make a vintage gift tag'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 42, "task": "dvc", "source": "hirest", "video": "hirest/pzvKbNsG-T8_226_326.mp4", "duration": 100.0, "tgt": [[0, 21], [21, 38], [38, 58], [58, 74], [74, 100], [100, 100]], "g": ["take paper collection put on top", "pain color on flowers", "put white patches on edges", "put greeting on inside cover", "put message on paper and send", "create calendar on another method"], "q": "You are given a video about 'make a vintage gift tag'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 43, "task": "dvc", "source": "hirest", "video": "hirest/LFufa6Nfc6Y_68_189.mp4", "duration": 121.0, "tgt": [[0, 4], [4, 19], [19, 29], [29, 41], [41, 49], [49, 64], [64, 84], [84, 121]], "g": ["snow look so good", "fourteen desert are there", "get to find a chopping", "set seven topping here", "put a jelly here", "arrange all jelly into show", "all can be yummy", "set all the candy here"], "q": "You are given a video about 'make snow cones'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 44, "task": "dvc", "source": "hirest", "video": "hirest/iKRdPAsOPPk_19_256.mp4", "duration": 237.0, "tgt": [[0, 5], [5, 13], [13, 23], [23, 63], [63, 75], [75, 98], [98, 163], [163, 237]], "g": ["fold sheet into cone shape", "use three different colors to stick", "tape all cones together", "pin plastic cover inside the cone", "put ices in the jar", "grind it well to make snow", "fill snow to cone sheet", "add some colors for topping"], "q": "You are given a video about 'make snow cones'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 45, "task": "dvc", "source": "hirest", "video": "hirest/ZROZTtnz2Jc_69_224.mp4", "duration": 155.0, "tgt": [[0, 6], [6, 11], [11, 47], [47, 69], [69, 72], [72, 74], [74, 89], [89, 100], [100, 106], [106, 108], [108, 128], [128, 131], [131, 150], [150, 155]], "g": ["cut the chocolate into half pieces", "melt white chocolate cookies in oven", "spread it on sheet", "design toys with colorful cookies", "bake it in oven", "cut the pieces with separate design", "take some skumfiduser marshmallow", "take some chocolate cookies for decoration", "transfer the cookies into plastic bag", "mash it well", "dip mash mellow with chocolate", "dip into the mashed cookies", "insert bamboo stick onto that", "pack it like a gift individually"], "q": "You are given a video about 'make a christmas gift for a teacher'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 46, "task": "dvc", "source": "hirest", "video": "hirest/iFMMxvqemww_105_226.mp4", "duration": 121.0, "tgt": [[0, 37], [37, 79], [79, 121]], "g": ["get a bottle of body spray", "tie the body spray with ribbon", "put body spray into paper bag"], "q": "You are given a video about 'make a christmas gift for a teacher'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 47, "task": "dvc", "source": "hirest", "video": "hirest/DBBJe_kdLDI_99_168.mp4", "duration": 69.0, "tgt": [[0, 10], [10, 23], [23, 33], [33, 49], [49, 56], [56, 69]], "g": ["use trim", "put stamps on top", "use gold paper", "make flowers for bin", "use pearls", "use sweet peppermint collection"], "q": "You are given a video about 'make a christmas gift for a teacher'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 48, "task": "dvc", "source": "hirest", "video": "hirest/ZujIOKsOnzw_27_133.mp4", "duration": 106.0, "tgt": [[0, 36], [36, 47], [47, 60], [60, 81], [81, 102], [102, 106]], "g": ["take an materials", "take designed chart and santa", "take tissue and candy on bottle", "take some glitters", "fold glitter on tissue", "put all candy tissue on bottle"], "q": "You are given a video about 'make a christmas gift for a teacher'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 49, "task": "dvc", "source": "hirest", "video": "hirest/gWR8Nsf_Z_g_61_221.mp4", "duration": 160.0, "tgt": [[0, 18], [18, 41], [41, 62], [62, 86], [86, 88], [88, 100], [100, 114], [114, 136], [136, 146], [146, 160]], "g": ["unscrew the top", "lift the top off", "empty the container", "use scoop bottle", "get resin", "put in container", "add water", "put filter back", "screw the top tight", "connect the taps"], "q": "You are given a video about 'clean resin'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 50, "task": "dvc", "source": "hirest", "video": "hirest/yH3p_oMRyns_144_498.mp4", "duration": 354.0, "tgt": [[0, 20], [20, 39], [39, 82], [82, 122], [122, 129], [129, 228], [228, 263], [263, 305], [305, 324], [324, 354]], "g": ["attach dahlia and garden roses", "add spray roses", "add greens", "add brunia", "put another dahlia", "add another rose", "put floral tape", "wrap the bouquet", "cut the ribbon", "put pins"], "q": "You are given a video about 'make a bouquet'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 51, "task": "dvc", "source": "hirest", "video": "hirest/oDRoXgINJAU_108_288.mp4", "duration": 180.0, "tgt": [[0, 13], [13, 25], [25, 50], [50, 70], [70, 76], [76, 89], [89, 99], [99, 120], [120, 140], [140, 148], [148, 168], [168, 180]], "g": ["boil water and add meat on", "fry vegetables using pan", "add carrot on pan", "take out meat", "add sous on boiled meat water", "add cabbage into water", "add carrot on it", "mix all well", "add sous on it", "add tomato and needed herbs", "cook it well", "serve on another bowl"], "q": "You are given a video about 'make borsh russian soup'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 52, "task": "dvc", "source": "hirest", "video": "hirest/sk9KApzKeLw_54_120.mp4", "duration": 66.0, "tgt": [[0, 6], [6, 12], [12, 14], [14, 22], [22, 66]], "g": ["hold the coin between the tips of your fingers", "reach your right hand over the coin to conceal it", "drop the coin into your left palm", "pull your right hand away", "open your empty right hand slowly"], "q": "You are given a video about 'do the 'french drop''. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 53, "task": "dvc", "source": "hirest", "video": "hirest/nN6ZOXoBjps_57_130.mp4", "duration": 73.0, "tgt": [[0, 72], [72, 73]], "g": ["rub off the dust and oxide", "dry immediately"], "q": "You are given a video about 'clean chrome'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 54, "task": "dvc", "source": "hirest", "video": "hirest/YY_JX65mWuw_67_154.mp4", "duration": 87.0, "tgt": [[0, 6], [6, 45], [45, 57], [57, 87]], "g": ["pour cleaning liquid in cloth", "scrub chrome with cloth", "take more liquid to clean more stain", "wipe it by dry cloth"], "q": "You are given a video about 'clean chrome'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 55, "task": "dvc", "source": "hirest", "video": "hirest/daclapuOQnE_73_157.mp4", "duration": 84.0, "tgt": [[0, 9], [9, 25], [25, 41], [41, 74], [74, 84]], "g": ["take scrub and spray water", "rub out the scratches", "use cloth to clean out", "spray water and clean with water", "use scrubber to clean it well"], "q": "You are given a video about 'clean chrome'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 56, "task": "dvc", "source": "hirest", "video": "hirest/qE5eodQCqqc_22_119.mp4", "duration": 97.0, "tgt": [[0, 16], [16, 76], [76, 97]], "g": ["take metal polish into cloth", "wiping down chrome gently", "clean by new cloth fully"], "q": "You are given a video about 'clean chrome'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 57, "task": "dvc", "source": "hirest", "video": "hirest/qPzWyc-XTQ4_37_283.mp4", "duration": 246.0, "tgt": [[0, 4], [4, 35], [35, 44], [44, 78], [78, 90], [90, 96], [96, 199], [199, 208], [208, 226], [226, 229], [229, 237], [237, 246]], "g": ["shake and spray it", "keep doing the process", "let it sit", "rub it around", "wash with water", "dry it off", "repeat the process", "polish it", "use sandpaper for roughness", "use water proof sandpaper", "polish it more", "use different sandpapers"], "q": "You are given a video about 'clean chrome'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 58, "task": "dvc", "source": "hirest", "video": "hirest/ymAgcwbWo38_28_47.mp4", "duration": 19.0, "tgt": [[0, 8], [8, 12], [12, 19]], "g": ["apply liquid on parts", "rub it out the dirty", "use cloth to clean out"], "q": "You are given a video about 'clean chrome'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 59, "task": "dvc", "source": "hirest", "video": "hirest/94Kea14_1I0_58_321.mp4", "duration": 263.0, "tgt": [[0, 10], [10, 21], [21, 56], [56, 263]], "g": ["mark and put lines on paper", "fold it properly", "fix it with tape", "add some color papers"], "q": "You are given a video about 'make a duct tape wallet'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 60, "task": "dvc", "source": "hirest", "video": "hirest/PiIfYPwuw7c_6_368.mp4", "duration": 362.0, "tgt": [[0, 4], [4, 11], [11, 27], [27, 33], [33, 42], [42, 61], [61, 75], [75, 85], [85, 104], [104, 122], [122, 146], [146, 154], [154, 204], [204, 219], [219, 230], [230, 233], [233, 286], [286, 350], [350, 362]], "g": ["leave all tape on borders", "trim it", "cut squares out", "fold the divider", "put it on", "cut a slit", "fold it over", "line the other on edge", "fold the tape", "trim the lower part", "fold the tape", "grab coin pouch", "attach the tape", "cut and fold the tape", "do same on other side", "fold the flap down", "fold rest of tape", "put velcro", "put little tape on edges"], "q": "You are given a video about 'make a duct tape wallet'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 61, "task": "dvc", "source": "hirest", "video": "hirest/AEkQeihliZM_29_95.mp4", "duration": 66.0, "tgt": [[0, 15], [15, 35], [35, 58], [58, 66]], "g": ["arrange wanted length on room", "using spray to spread water", "using cloth to clean out floor", "use machine to clean out"], "q": "You are given a video about 'apply diatomaceous earth to carpet'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 62, "task": "dvc", "source": "hirest", "video": "hirest/n_h7GGgUhDM_73_270.mp4", "duration": 197.0, "tgt": [[0, 20], [20, 32], [32, 82], [82, 137], [137, 176], [176, 197]], "g": ["pick the plant", "select a container with drainage", "fill up with water", "put it on sunny side", "put water not too much", "dry out some time"], "q": "You are given a video about 'grow fresh herbs'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 63, "task": "dvc", "source": "hirest", "video": "hirest/yWpSs787IRI_55_220.mp4", "duration": 165.0, "tgt": [[0, 14], [14, 34], [34, 49], [49, 66], [66, 94], [94, 114], [114, 137], [137, 156], [156, 165]], "g": ["use elastic to create belly", "take pins and thread", "put elastic on dress back", "put three inch elastic on side", "stich elastic using threads and pins", "put stich on other side elastic", "put stich on back ward direction", "tight out elastic on sides", "wear dress as comfort"], "q": "You are given a video about 'make a full belly dance skirt'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 64, "task": "dvc", "source": "hirest", "video": "hirest/Dz_BZ48BhsA_194_535.mp4", "duration": 341.0, "tgt": [[0, 7], [7, 11], [11, 19], [19, 43], [43, 50], [50, 92], [92, 119], [119, 145], [145, 148], [148, 177], [177, 181], [181, 185], [185, 226], [226, 341]], "g": ["measure the size", "leave extra on sides", "cut the edging", "do same with other", "make into regular bra", "sew the edging", "put on bra", "sew bra strap inside", "hook it all", "adjust and cut it", "make thin belt first", "hook it with hip scarf", "embellish the scarf", "add rhinestones"], "q": "You are given a video about 'make a full belly dance skirt'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 65, "task": "dvc", "source": "hirest", "video": "hirest/qGCVZTvjMFw_29_259.mp4", "duration": 230.0, "tgt": [[0, 59], [59, 117], [117, 209], [209, 230]], "g": ["cut the material", "cut like mermaid dress", "mark into different pieces", "stich it"], "q": "You are given a video about 'make a full belly dance skirt'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 66, "task": "dvc", "source": "hirest", "video": "hirest/IX86Gre7d2Y_49_129.mp4", "duration": 80.0, "tgt": [[0, 1], [1, 9], [9, 21], [21, 26], [26, 27], [27, 48], [48, 51], [51, 80]], "g": ["turn skirt inside out", "locate the seams", "draw lines", "sew along the lines", "do back for no side seam", "make dart and stitch", "take piece of elastic", "stitch it at back of skirt"], "q": "You are given a video about 'make a full belly dance skirt'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 67, "task": "dvc", "source": "hirest", "video": "hirest/WIHPzX3zLNo_53_99.mp4", "duration": 46.0, "tgt": [[0, 14], [14, 20], [20, 46]], "g": ["peel the duct tape into two", "joint tape one by one", "repeat the step upto make bowl"], "q": "You are given a video about 'make a duct tape bowl'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 68, "task": "dvc", "source": "hirest", "video": "hirest/LH6B1Ye8hYw_17_80.mp4", "duration": 63.0, "tgt": [[0, 13], [13, 14], [14, 18], [18, 50], [50, 53], [53, 63]], "g": ["take one cup of heavy cream", "add two tsp of buttermilk", "wrap it with plastic cover", "keep it room temperature for eight hours", "transfer into other bowl", "keep it in fridge"], "q": "You are given a video about 'make creme fraiche'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 69, "task": "dvc", "source": "hirest", "video": "hirest/8_Mg-mOvvMw_67_191.mp4", "duration": 124.0, "tgt": [[0, 13], [13, 28], [28, 29], [29, 40], [40, 48], [48, 68], [68, 109], [109, 124]], "g": ["cut roma tomatoes", "spray oil , salt and bake", "cut beetroot and garlic", "add tahini and lemon juice", "add canon oil,salt and mash it", "prepare toppings with parsley", "cut the bread into slices and toast", "assemble everything with omelette"], "q": "You are given a video about 'create a brunch menu'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 70, "task": "dvc", "source": "hirest", "video": "hirest/6C0mL9n9Nko_11_116.mp4", "duration": 105.0, "tgt": [[0, 29], [29, 48], [48, 59], [59, 74], [74, 98], [98, 105]], "g": ["have a needed things", "using a clay design", "another clay", "start pour gycerlin", "remove outer layer of clay", "using chain measure the band size"], "q": "You are given a video about 'make a fairy headband'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 71, "task": "dvc", "source": "hirest", "video": "hirest/vxDNQjk75dU_56_168.mp4", "duration": 112.0, "tgt": [[0, 26], [26, 43], [43, 50], [50, 66], [66, 73], [73, 77], [77, 82], [82, 91], [91, 96], [96, 101], [101, 104], [104, 112]], "g": ["put almond flour in bowl", "add confectioners", "add chocolate chips", "add salt", "add peanut butter", "add melted coconut oil", "add vanilla extract", "mix it", "make a cookie dough", "scoop out the cookie dough", "place on baking sheet", "put in fridge"], "q": "You are given a video about 'make peanut butter cup cookie bites'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 72, "task": "dvc", "source": "hirest", "video": "hirest/WqVQSC5zMBw_20_30.mp4", "duration": 10.0, "tgt": [[0, 5], [5, 10]], "g": ["put mashed potato into freezer", "using bag to store as well"], "q": "You are given a video about 'freeze mashed potatoes'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 73, "task": "dvc", "source": "hirest", "video": "hirest/e76KWpbPdfY_107_465.mp4", "duration": 358.0, "tgt": [[0, 36], [36, 70], [70, 116], [116, 136], [136, 165], [165, 174], [174, 180], [180, 196], [196, 220], [220, 227], [227, 229], [229, 240], [240, 255], [255, 258], [258, 280], [280, 315], [315, 321], [321, 325], [325, 337], [337, 351], [351, 358]], "g": ["take large piece", "feel vertical bars", "take small hinge", "feel the top bar", "put small piece over large", "do both same time", "grab corner tightly", "put pin on right", "stretch the large piece", "put bottom bar on hook", "hold it there", "stretch the hinges", "use blade", "get key cap", "push it on sides", "turn the large key", "put wire on bottom hinges", "turn key over", "open up the wire", "put wire under", "turn key over and push"], "q": "You are given a video about 'fix sticky keyboard keys'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}, {"version": 1.0, "idx": 74, "task": "dvc", "source": "hirest", "video": "hirest/L8kJRfPMAEY_35_115.mp4", "duration": 80.0, "tgt": [[0, 1], [1, 3], [3, 8], [8, 12], [12, 14], [14, 36], [36, 45], [45, 47], [47, 80]], "g": ["pour cereal", "add rice krispies and fruity pebbles", "add marshmallows", "mix it around", "put it in the microwave", "watch the heat", "use hot gloves to take out", "put treats in the dish", "mold the treats"], "q": "You are given a video about 'make rice krispies paint brush treats'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'. You must represent start and end times in seconds. For example: '60 - 62 seconds, add salt and pepper in eggs. 62 - 65 seconds, add cayenne pepper and bacon.'."}]
annotations/txt/dvc_youcook2.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/txt/eca_charades_sta.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/txt/epm_eg4d_nlq.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/txt/evs_summe.json ADDED
@@ -0,0 +1 @@
 
 
1
+ [{"version": 1.0, "idx": 0, "task": "evs", "source": "summe", "video": "summe/saving_dolphines.mp4", "duration": 223.333, "tgt": [[3.37, 5.172], [5.939, 6.907], [7.708, 9.643], [23.023, 26.427], [27.127, 33.601], [36.47, 43.911], [49.684, 52.386], [57.058, 61.295], [65.7, 68.336], [76.677, 80.782], [110.111, 112.981], [134.302, 138.073], [201.637, 202.004]], "q": "You are given a video about saving dolphines. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 1, "task": "evs", "source": "summe", "video": "summe/bearpark_climbing.mp4", "duration": 134.0, "tgt": [[3.4, 4.48], [9.6, 12.6], [48.081, 48.361], [48.641, 50.481], [76.842, 80.322], [83.962, 85.922], [85.962, 87.162], [101.322, 102.442], [103.482, 106.642], [106.682, 107.082], [107.562, 107.602], [109.362, 109.722], [109.962, 110.322], [110.922, 112.003], [112.043, 112.563], [124.443, 125.963]], "q": "You are given a video about bearpark climbing. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 2, "task": "evs", "source": "summe", "video": "summe/playing_ball.mp4", "duration": 104.333, "tgt": [[1.267, 4.267], [5.767, 5.967], [11.267, 11.467], [11.501, 12.701], [12.867, 13.367], [13.867, 22.334], [29.435, 31.202], [53.203, 53.869], [53.903, 57.536], [86.671, 87.704]], "q": "You are given a video about playing ball. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 3, "task": "evs", "source": "summe", "video": "summe/car_railcrossing.mp4", "duration": 169.667, "tgt": [[3.337, 4.271], [10.444, 12.013], [13.347, 13.715], [14.015, 26.194], [78.45, 86.625], [86.725, 87.893], [112.386, 113.153], [157.166, 158.334], [158.635, 161.237]], "q": "You are given a video about car railcrossing. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 4, "task": "evs", "source": "summe", "video": "summe/st_maarten_landing.mp4", "duration": 70.333, "tgt": [[25.32, 28.52], [48.2, 55.88]], "q": "You are given a video about st maarten landing. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 5, "task": "evs", "source": "summe", "video": "summe/excavators_river_crossing.mp4", "duration": 389.167, "tgt": [[2.52, 9.72], [10.08, 12.96], [27.0, 29.36], [35.48, 50.521], [62.641, 63.201], [69.201, 83.441], [93.681, 101.921], [102.801, 114.241], [114.641, 115.041], [119.241, 121.801], [122.401, 123.201], [377.844, 378.284]], "q": "You are given a video about excavators river crossing. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 6, "task": "evs", "source": "summe", "video": "summe/uncut_evening_flight.mp4", "duration": 323.0, "tgt": [[0.0, 15.049], [17.285, 18.453], [65.835, 66.536], [68.972, 69.039], [69.873, 72.075], [88.726, 95.199], [111.016, 117.156], [163.771, 170.445], [286.566, 287.467], [287.567, 288.735], [289.169, 290.737], [307.054, 320.802]], "q": "You are given a video about uncut evening flight. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 7, "task": "evs", "source": "summe", "video": "summe/paluma_jump.mp4", "duration": 86.167, "tgt": [[40.044, 45.083], [59.966, 66.706], [78.619, 79.454], [81.623, 82.457]], "q": "You are given a video about paluma jump. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 8, "task": "evs", "source": "summe", "video": "summe/statue_of_liberty.mp4", "duration": 154.833, "tgt": [[24.174, 35.778], [41.418, 42.587], [43.513, 43.675], [45.971, 47.019], [47.059, 49.597], [57.333, 57.655], [97.422, 100.887], [113.86, 117.204], [117.285, 119.138], [121.757, 121.999], [122.2, 122.684], [132.796, 135.939]], "q": "You are given a video about statue of liberty. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 9, "task": "evs", "source": "summe", "video": "summe/air_force_one.mp4", "duration": 180.167, "tgt": [[56.964, 57.204], [66.804, 67.164], [68.125, 68.925], [70.405, 85.926], [155.77, 171.531], [173.772, 174.412]], "q": "You are given a video about air force one. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 10, "task": "evs", "source": "summe", "video": "summe/eiffel_tower.mp4", "duration": 199.167, "tgt": [[3.095, 6.713], [30.712, 31.516], [47.676, 56.037], [79.433, 81.644], [100.055, 101.1], [101.623, 104.356], [113.723, 114.446], [115.572, 119.592], [126.345, 131.008], [134.305, 138.968], [195.447, 198.623]], "q": "You are given a video about eiffel tower. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 11, "task": "evs", "source": "summe", "video": "summe/kids_playing_in_leaves.mp4", "duration": 106.667, "tgt": [[2.336, 14.683], [15.484, 20.055], [30.967, 32.836], [42.313, 44.482], [53.492, 54.059], [68.175, 68.308], [71.679, 73.914], [91.434, 92.068], [93.035, 94.137], [96.773, 97.741]], "q": "You are given a video about kids playing in leaves. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 12, "task": "evs", "source": "summe", "video": "summe/car_over_camera.mp4", "duration": 146.5, "tgt": [[36.743, 37.377], [38.011, 38.511], [56.766, 59.269], [60.77, 64.808], [65.609, 65.776], [85.799, 92.941], [92.974, 93.742], [101.551, 102.419], [102.852, 107.591], [108.392, 112.03], [140.863, 143.766]], "q": "You are given a video about car over camera. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 13, "task": "evs", "source": "summe", "video": "summe/fire_domino.mp4", "duration": 54.0, "tgt": [[9.132, 13.206], [13.274, 13.375], [18.841, 19.86], [23.865, 24.612], [25.461, 25.733], [44.574, 47.357]], "q": "You are given a video about fire domino. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 14, "task": "evs", "source": "summe", "video": "summe/cockpit_landing.mp4", "duration": 302.167, "tgt": [[5.414, 6.918], [38.298, 41.673], [72.652, 83.179], [111.05, 120.574], [240.112, 244.223], [244.724, 247.364], [276.505, 277.608], [281.05, 284.358], [285.695, 301.034]], "q": "You are given a video about cockpit landing. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 15, "task": "evs", "source": "summe", "video": "summe/valparaiso_downhill.mp4", "duration": 173.167, "tgt": [[10.278, 16.284], [23.525, 24.559], [28.697, 29.932], [31.4, 31.467], [34.136, 35.738], [35.938, 35.971], [37.273, 37.64], [39.475, 42.145], [66.504, 69.14], [69.874, 70.141], [89.028, 92.431], [96.269, 97.67], [127.769, 129.737], [136.044, 137.779], [160.036, 166.743], [166.843, 167.01]], "q": "You are given a video about valparaiso downhill. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 16, "task": "evs", "source": "summe", "video": "summe/jumps.mp4", "duration": 38.333, "tgt": [[2.069, 3.517], [9.062, 9.352], [9.683, 10.635], [12.083, 15.601], [24.456, 24.704]], "q": "You are given a video about jumps. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 17, "task": "evs", "source": "summe", "video": "summe/base_jumping.mp4", "duration": 158.167, "tgt": [[4.232, 6.113], [27.037, 33.049], [42.89, 49.204], [50.38, 50.581], [50.615, 53.503], [68.92, 69.793], [81.817, 86.855], [95.99, 96.192], [96.293, 97.502], [98.61, 100.961], [122.994, 128.939], [135.555, 136.16], [136.429, 136.731]], "q": "You are given a video about base jumping. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 18, "task": "evs", "source": "summe", "video": "summe/bike_polo.mp4", "duration": 102.5, "tgt": [[21.619, 26.09], [28.041, 31.974], [62.2, 62.503], [63.142, 63.444], [70.269, 75.615], [85.702, 88.123]], "q": "You are given a video about bike polo. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 19, "task": "evs", "source": "summe", "video": "summe/notre_dame.mp4", "duration": 192.333, "tgt": [[4.459, 9.668], [11.002, 11.502], [17.586, 17.92], [21.754, 22.337], [23.212, 24.254], [24.629, 25.088], [89.098, 95.308], [96.808, 98.016], [118.728, 118.811], [123.187, 131.23], [137.731, 140.398], [142.357, 142.607], [143.024, 144.524], [176.696, 180.53]], "q": "You are given a video about notre dame. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 20, "task": "evs", "source": "summe", "video": "summe/playing_on_water_slide.mp4", "duration": 102.667, "tgt": [[6.975, 8.51], [10.913, 10.98], [11.08, 11.113], [15.018, 15.285], [28.634, 34.407], [34.474, 34.674], [34.741, 36.543], [45.053, 48.057], [52.128, 52.896], [64.142, 65.41], [91.541, 91.875], [92.042, 99.717]], "q": "You are given a video about playing on water slide. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 21, "task": "evs", "source": "summe", "video": "summe/paintball.mp4", "duration": 254.667, "tgt": [[7.088, 15.183], [36.153, 40.683], [55.278, 63.205], [97.68, 106.32], [230.507, 236.378], [238.224, 245.06]], "q": "You are given a video about paintball. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 22, "task": "evs", "source": "summe", "video": "summe/bus_in_rock_tunnel.mp4", "duration": 171.5, "tgt": [[47.114, 52.619], [53.053, 55.956], [67.667, 69.403], [69.569, 70.537], [75.976, 76.343], [84.184, 84.518], [85.318, 85.385], [85.819, 86.119], [88.054, 97.464], [125.759, 126.226], [139.472, 140.106], [140.173, 140.507], [146.146, 147.914], [149.015, 149.082], [166.332, 168.735]], "q": "You are given a video about bus in rock tunnel. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 23, "task": "evs", "source": "summe", "video": "summe/scuba.mp4", "duration": 74.333, "tgt": [[1.6, 2.333], [6.533, 6.9], [15.7, 19.167], [19.7, 21.567], [36.833, 40.4], [55.266, 57.833], [62.3, 62.633], [63.8, 66.1], [71.933, 72.2], [72.466, 73.866]], "q": "You are given a video about scuba. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 24, "task": "evs", "source": "summe", "video": "summe/cooking.mp4", "duration": 86.167, "tgt": [[30.968, 31.574], [32.92, 39.181], [51.838, 54.194], [68.13, 69.005], [69.072, 74.054]], "q": "You are given a video about cooking. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}]
annotations/txt/evs_tvsum.json ADDED
@@ -0,0 +1 @@
 
 
1
+ [{"version": 1.0, "idx": 0, "task": "evs", "source": "tvsum", "video": "tvsum/akI8YFjEmUw.mp4", "duration": 133.333, "tgt": [[32, 40], [44, 46], [66, 68], [110, 112], [122, 126], [128, 130]], "q": "You are given a video about getting vehicle unstuck. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 1, "task": "evs", "source": "tvsum", "video": "tvsum/J0nA4VgnoCo.mp4", "duration": 584.667, "tgt": [[94, 100], [110, 112], [240, 246], [248, 252], [266, 276], [290, 308], [320, 334], [336, 344], [346, 350], [352, 356], [364, 370], [374, 376], [410, 414]], "q": "You are given a video about changing vehicle tire. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 2, "task": "evs", "source": "tvsum", "video": "tvsum/kLxoNp-UchI.mp4", "duration": 130.0, "tgt": [[8, 14], [26, 28], [50, 52], [84, 86], [96, 98], [100, 102], [106, 108], [118, 120]], "q": "You are given a video about dog show. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 3, "task": "evs", "source": "tvsum", "video": "tvsum/WxtbjNsCQ8A.mp4", "duration": 265.5, "tgt": [[66, 76], [82, 84], [92, 100], [122, 128], [140, 142], [172, 174], [222, 224], [232, 240]], "q": "You are given a video about bee keeping. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 4, "task": "evs", "source": "tvsum", "video": "tvsum/HT5vyqe0Xaw.mp4", "duration": 322.667, "tgt": [[38, 40], [42, 50], [58, 60], [64, 68], [82, 86], [110, 112], [120, 122], [148, 152], [158, 166], [168, 170], [236, 238], [258, 260], [276, 278], [280, 282], [284, 286], [298, 302]], "q": "You are given a video about getting vehicle unstuck. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 5, "task": "evs", "source": "tvsum", "video": "tvsum/gzDbaEs1Rlg.mp4", "duration": 288.333, "tgt": [[34, 36], [70, 72], [76, 78], [104, 110], [114, 116], [120, 122], [136, 140], [144, 150], [154, 164], [172, 176], [204, 206], [222, 224]], "q": "You are given a video about changing vehicle tire. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 6, "task": "evs", "source": "tvsum", "video": "tvsum/EE-bNr36nyA.mp4", "duration": 98.167, "tgt": [[30, 36], [42, 44], [46, 50], [80, 82], [86, 88]], "q": "You are given a video about bee keeping. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 7, "task": "evs", "source": "tvsum", "video": "tvsum/eQu1rNs0an0.mp4", "duration": 164.5, "tgt": [[4, 6], [12, 14], [22, 24], [48, 52], [64, 72], [106, 112], [136, 140]], "q": "You are given a video about attempting bike tricks. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 8, "task": "evs", "source": "tvsum", "video": "tvsum/oDXZc0tZe04.mp4", "duration": 380.833, "tgt": [[2, 4], [92, 96], [98, 100], [112, 118], [152, 154], [156, 158], [200, 210], [250, 256], [294, 300], [302, 312], [334, 340], [364, 366], [368, 370]], "q": "You are given a video about bee keeping. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 9, "task": "evs", "source": "tvsum", "video": "tvsum/Hl-__g2gn_A.mp4", "duration": 243.833, "tgt": [[8, 10], [62, 64], [72, 74], [76, 80], [84, 86], [110, 114], [128, 132], [166, 174], [186, 188], [206, 208], [224, 230]], "q": "You are given a video about making sandwich. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 10, "task": "evs", "source": "tvsum", "video": "tvsum/WG0MBPpPC6I.mp4", "duration": 397.667, "tgt": [[8, 10], [52, 54], [106, 108], [126, 128], [168, 170], [180, 182], [190, 194], [222, 224], [242, 244], [256, 260], [266, 270], [274, 278], [284, 288], [300, 302], [324, 330], [338, 348], [352, 354], [356, 358], [366, 368]], "q": "You are given a video about making sandwich. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 11, "task": "evs", "source": "tvsum", "video": "tvsum/GsAD1KT1xo8.mp4", "duration": 145.333, "tgt": [[76, 80], [86, 90], [92, 96], [114, 118], [122, 124], [128, 130], [136, 138]], "q": "You are given a video about parkour. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 12, "task": "evs", "source": "tvsum", "video": "tvsum/xxdtq8mxegs.mp4", "duration": 144.167, "tgt": [[70, 88], [90, 92], [102, 104]], "q": "You are given a video about grooming an animal. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 13, "task": "evs", "source": "tvsum", "video": "tvsum/NyBmCxDoHJU.mp4", "duration": 189.667, "tgt": [[20, 24], [40, 44], [60, 62], [66, 70], [106, 108], [114, 116], [118, 122], [128, 130], [136, 142], [176, 178]], "q": "You are given a video about dog show. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 14, "task": "evs", "source": "tvsum", "video": "tvsum/Bhxk-O1Y7Ho.mp4", "duration": 450.833, "tgt": [[0, 2], [74, 82], [102, 104], [108, 112], [114, 122], [160, 170], [174, 176], [268, 270], [272, 274], [276, 278], [286, 288], [292, 296], [298, 300], [326, 328], [330, 334], [336, 340], [434, 440], [442, 450]], "q": "You are given a video about grooming an animal. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 15, "task": "evs", "source": "tvsum", "video": "tvsum/_xMr-HKMfVA.mp4", "duration": 148.833, "tgt": [[18, 20], [60, 62], [80, 90], [94, 100], [118, 122]], "q": "You are given a video about flash mob gathering. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 16, "task": "evs", "source": "tvsum", "video": "tvsum/i3wAGJaaktw.mp4", "duration": 156.833, "tgt": [[4, 8], [10, 12], [28, 32], [56, 58], [64, 66], [88, 100]], "q": "You are given a video about grooming an animal. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 17, "task": "evs", "source": "tvsum", "video": "tvsum/JKpqYvAdIsw.mp4", "duration": 152.0, "tgt": [[18, 20], [40, 42], [48, 50], [52, 60], [108, 110], [124, 126], [128, 132], [140, 142]], "q": "You are given a video about flash mob gathering. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 18, "task": "evs", "source": "tvsum", "video": "tvsum/byxOvuiIJV0.mp4", "duration": 154.5, "tgt": [[30, 34], [36, 40], [48, 50], [66, 68], [78, 80], [98, 100], [102, 108], [128, 132]], "q": "You are given a video about flash mob gathering. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 19, "task": "evs", "source": "tvsum", "video": "tvsum/RBCABdttQmI.mp4", "duration": 364.333, "tgt": [[6, 10], [32, 36], [84, 86], [88, 90], [124, 126], [216, 218], [222, 236], [266, 268], [272, 276], [310, 318], [326, 328], [336, 342], [346, 348], [352, 354], [356, 358]], "q": "You are given a video about parade. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 20, "task": "evs", "source": "tvsum", "video": "tvsum/XzYM3PfTM4w.mp4", "duration": 111.0, "tgt": [[20, 24], [32, 34], [48, 58], [68, 70]], "q": "You are given a video about changing vehicle tire. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 21, "task": "evs", "source": "tvsum", "video": "tvsum/0tmA_C6XwfM.mp4", "duration": 141.333, "tgt": [[28, 30], [52, 78], [106, 108]], "q": "You are given a video about grooming an animal. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 22, "task": "evs", "source": "tvsum", "video": "tvsum/qqR6AEXwxoQ.mp4", "duration": 269.333, "tgt": [[56, 62], [80, 86], [118, 120], [148, 164], [166, 174], [240, 246]], "q": "You are given a video about attempting bike tricks. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 23, "task": "evs", "source": "tvsum", "video": "tvsum/vdmoEJ5YbrQ.mp4", "duration": 329.333, "tgt": [[32, 34], [64, 66], [84, 86], [96, 98], [100, 112], [126, 128], [132, 138], [140, 156], [178, 180], [184, 188]], "q": "You are given a video about getting vehicle unstuck. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 24, "task": "evs", "source": "tvsum", "video": "tvsum/LRw_obCPUt0.mp4", "duration": 260.333, "tgt": [[4, 8], [58, 68], [118, 132], [136, 138], [156, 160], [234, 242]], "q": "You are given a video about making sandwich. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 25, "task": "evs", "source": "tvsum", "video": "tvsum/uGu_10sucQo.mp4", "duration": 167.0, "tgt": [[30, 32], [36, 38], [40, 42], [64, 66], [72, 74], [76, 78], [96, 108], [112, 120], [124, 126]], "q": "You are given a video about bee keeping. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 26, "task": "evs", "source": "tvsum", "video": "tvsum/4wU_LUjG5Ic.mp4", "duration": 167.0, "tgt": [[8, 14], [16, 22], [52, 56], [58, 60], [102, 108], [138, 140]], "q": "You are given a video about parade. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 27, "task": "evs", "source": "tvsum", "video": "tvsum/98MoyGZKHXc.mp4", "duration": 187.5, "tgt": [[28, 30], [64, 70], [76, 86], [102, 108], [116, 118], [128, 130]], "q": "You are given a video about changing vehicle tire. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 28, "task": "evs", "source": "tvsum", "video": "tvsum/XkqCExn6_Us.mp4", "duration": 187.833, "tgt": [[2, 4], [22, 24], [26, 28], [56, 58], [64, 66], [74, 84], [86, 88], [92, 94], [100, 102], [106, 108]], "q": "You are given a video about parkour. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 29, "task": "evs", "source": "tvsum", "video": "tvsum/Se3oxnaPsz0.mp4", "duration": 138.833, "tgt": [[42, 44], [52, 56], [76, 78], [86, 90], [102, 106], [110, 116]], "q": "You are given a video about bee keeping. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 30, "task": "evs", "source": "tvsum", "video": "tvsum/sTEELN-vY30.mp4", "duration": 149.0, "tgt": [[0, 14], [66, 74], [118, 122], [130, 132]], "q": "You are given a video about getting vehicle unstuck. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 31, "task": "evs", "source": "tvsum", "video": "tvsum/JgHubY5Vw3Y.mp4", "duration": 143.667, "tgt": [[12, 14], [80, 82], [94, 112]], "q": "You are given a video about attempting bike tricks. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 32, "task": "evs", "source": "tvsum", "video": "tvsum/3eYKfiOEJNs.mp4", "duration": 194.167, "tgt": [[50, 58], [60, 62], [64, 66], [76, 78], [82, 92], [170, 174], [176, 178]], "q": "You are given a video about grooming an animal. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 33, "task": "evs", "source": "tvsum", "video": "tvsum/xwqBXPGE9pQ.mp4", "duration": 233.833, "tgt": [[20, 22], [30, 34], [58, 62], [70, 74], [76, 88], [106, 108], [122, 130], [166, 172], [174, 176], [224, 226]], "q": "You are given a video about getting vehicle unstuck. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 34, "task": "evs", "source": "tvsum", "video": "tvsum/VuWGsYPqAX8.mp4", "duration": 216.5, "tgt": [[62, 64], [84, 88], [92, 94], [112, 116], [118, 120], [122, 126], [130, 132], [142, 144], [178, 180], [182, 184], [194, 198], [204, 206], [212, 214]], "q": "You are given a video about flash mob gathering. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 35, "task": "evs", "source": "tvsum", "video": "tvsum/jcoYJXDG9sw.mp4", "duration": 199.167, "tgt": [[90, 94], [96, 108], [122, 126], [158, 168]], "q": "You are given a video about dog show. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 36, "task": "evs", "source": "tvsum", "video": "tvsum/z_6gVvQb2d0.mp4", "duration": 276.333, "tgt": [[132, 136], [138, 144], [152, 154], [160, 162], [168, 170], [172, 174], [190, 192], [196, 198], [202, 204], [212, 220], [224, 228], [232, 234], [236, 238], [240, 242], [246, 248]], "q": "You are given a video about parade. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 37, "task": "evs", "source": "tvsum", "video": "tvsum/-esJrBWj2d8.mp4", "duration": 230.333, "tgt": [[18, 20], [42, 54], [56, 58], [64, 80], [110, 114], [116, 118], [126, 128]], "q": "You are given a video about dog show. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 38, "task": "evs", "source": "tvsum", "video": "tvsum/xmEERLqJ2kU.mp4", "duration": 446.0, "tgt": [[230, 232], [234, 236], [240, 244], [248, 250], [252, 258], [260, 262], [266, 268], [270, 272], [278, 280], [326, 340], [356, 358], [374, 394], [396, 398], [400, 404], [410, 414]], "q": "You are given a video about flash mob gathering. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 39, "task": "evs", "source": "tvsum", "video": "tvsum/b626MiF1ew4.mp4", "duration": 235.833, "tgt": [[0, 2], [18, 20], [48, 50], [58, 60], [62, 64], [72, 76], [80, 82], [112, 114], [124, 126], [128, 130], [152, 154], [156, 158], [192, 194], [208, 210], [212, 214], [218, 222]], "q": "You are given a video about parkour. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 40, "task": "evs", "source": "tvsum", "video": "tvsum/EYqVtI9YWJA.mp4", "duration": 198.167, "tgt": [[98, 100], [114, 116], [138, 140], [144, 146], [160, 174], [176, 182], [188, 190]], "q": "You are given a video about attempting bike tricks. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 41, "task": "evs", "source": "tvsum", "video": "tvsum/iVt07TCkFM0.mp4", "duration": 104.333, "tgt": [[8, 10], [44, 46], [52, 56], [58, 60], [66, 70], [72, 74]], "q": "You are given a video about attempting bike tricks. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 42, "task": "evs", "source": "tvsum", "video": "tvsum/fWutDQy1nnY.mp4", "duration": 585.833, "tgt": [[90, 92], [150, 154], [156, 158], [188, 194], [196, 216], [224, 228], [236, 238], [242, 244], [254, 256], [270, 274], [276, 280], [282, 290], [296, 298], [338, 340], [344, 346], [348, 354], [356, 360], [364, 366], [368, 374], [382, 384], [386, 388], [412, 416]], "q": "You are given a video about parade. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 43, "task": "evs", "source": "tvsum", "video": "tvsum/37rzWOQsNIw.mp4", "duration": 191.667, "tgt": [[0, 6], [16, 22], [62, 64], [66, 72], [74, 82], [154, 158], [162, 164]], "q": "You are given a video about making sandwich. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 44, "task": "evs", "source": "tvsum", "video": "tvsum/91IHQYk1IQM.mp4", "duration": 110.5, "tgt": [[18, 20], [38, 40], [42, 46], [52, 54], [62, 66], [82, 84], [86, 88]], "q": "You are given a video about parade. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 45, "task": "evs", "source": "tvsum", "video": "tvsum/cjibtmSLxQ4.mp4", "duration": 647.5, "tgt": [[8, 10], [30, 32], [38, 46], [60, 64], [74, 76], [78, 84], [122, 126], [146, 148], [150, 152], [172, 180], [182, 184], [190, 192], [198, 202], [258, 264], [272, 276], [322, 324], [380, 382], [384, 390], [518, 522], [524, 526], [528, 532], [534, 538], [578, 582], [584, 586], [588, 594], [606, 608], [618, 622], [628, 630]], "q": "You are given a video about parkour. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 46, "task": "evs", "source": "tvsum", "video": "tvsum/PJrm840pAUI.mp4", "duration": 274.167, "tgt": [[2, 4], [28, 30], [32, 34], [42, 48], [114, 116], [132, 134], [166, 168], [174, 184], [192, 194], [196, 202], [246, 248], [250, 252], [254, 262]], "q": "You are given a video about parkour. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 47, "task": "evs", "source": "tvsum", "video": "tvsum/Yi4Ij2NM7U4.mp4", "duration": 405.5, "tgt": [[58, 62], [124, 126], [130, 132], [134, 140], [142, 144], [156, 162], [166, 176], [236, 238], [240, 242], [254, 270], [294, 304], [306, 308], [334, 336], [370, 378]], "q": "You are given a video about making sandwich. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 48, "task": "evs", "source": "tvsum", "video": "tvsum/AwmHb44_ouw.mp4", "duration": 353.667, "tgt": [[0, 2], [94, 98], [102, 108], [120, 126], [138, 140], [142, 144], [152, 154], [156, 160], [164, 166], [172, 174], [188, 190], [192, 194], [198, 202], [208, 210], [222, 224], [256, 258], [260, 262], [274, 276], [284, 286], [300, 302]], "q": "You are given a video about changing vehicle tire. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}, {"version": 1.0, "idx": 49, "task": "evs", "source": "tvsum", "video": "tvsum/E11zDS9XGzg.mp4", "duration": 510.167, "tgt": [[40, 44], [90, 94], [110, 114], [130, 134], [158, 166], [168, 172], [174, 176], [178, 180], [242, 252], [260, 266], [268, 272], [288, 292], [298, 300], [304, 308], [318, 324], [378, 380], [388, 392], [422, 424], [440, 450], [456, 458]], "q": "You are given a video about dog show. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The summary locates in 5.2 - 7.5, 9.4 - 12.3, 16.9 - 18.2, and 21.8 - 25.4 seconds'."}]
annotations/txt/gvq_qa_ego4d.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/txt/rar_perception_test.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/txt/rvq_star.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/txt/slc_cross_task.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/txt/slc_ht_step.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/txt/tal_perception_test.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/txt/tal_thumos14.json ADDED
@@ -0,0 +1 @@
 
 
1
+ [{"version": 1.0, "idx": 0, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000324.mp4", "duration": 149.0, "tgt": [[49.2, 53.5], [116.7, 122.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 1, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000664.mp4", "duration": 86.0, "tgt": [[1.3, 4.8], [22.3, 25.3], [56.8, 59.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 2, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000714.mp4", "duration": 179.5, "tgt": [[26.2, 29.9], [37.0, 39.1], [141.4, 144.7], [150.1, 152.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 3, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000767.mp4", "duration": 77.167, "tgt": [[10.9, 12.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 4, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000873.mp4", "duration": 34.833, "tgt": [[6.3, 9.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 5, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000946.mp4", "duration": 27.833, "tgt": [[0.9, 3.5], [13.1, 17.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 6, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000964.mp4", "duration": 63.333, "tgt": [[7.5, 15.3], [42.5, 47.3], [52.2, 56.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 7, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001038.mp4", "duration": 121.167, "tgt": [[50.6, 54.2], [110.5, 114.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 8, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001182.mp4", "duration": 119.333, "tgt": [[1.3, 3.7], [78.1, 81.2], [89.9, 92.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 9, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001447.mp4", "duration": 229.833, "tgt": [[172.7, 175.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 10, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000006.mp4", "duration": 67.0, "tgt": [[18.8, 57.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 11, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000211.mp4", "duration": 98.833, "tgt": [[8.2, 10.1], [10.8, 12.5], [13.1, 14.8], [17.3, 19.5], [30.5, 40.1], [43.0, 44.2], [47.6, 49.3], [50.8, 52.5], [55.3, 62.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 12, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000426.mp4", "duration": 153.0, "tgt": [[27.1, 29.4], [34.9, 35.9], [53.1, 54.1], [55.6, 58.1], [63.2, 64.6], [78.9, 80.3], [84.3, 86.3], [100.1, 101.3], [144.1, 145.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 13, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000450.mp4", "duration": 28.333, "tgt": [[10.9, 13.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 14, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000615.mp4", "duration": 206.5, "tgt": [[10.0, 16.6], [52.6, 54.6], [111.9, 119.0], [128.4, 134.8], [171.4, 173.9], [184.0, 188.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 15, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000762.mp4", "duration": 133.0, "tgt": [[3.0, 4.8], [13.3, 15.8], [26.8, 29.1], [47.9, 50.1], [56.5, 58.0], [74.4, 76.8], [83.6, 85.8], [98.3, 101.6], [114.2, 118.4], [129.7, 132.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 16, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001078.mp4", "duration": 97.5, "tgt": [[0.5, 2.2], [4.6, 9.0], [15.3, 17.0], [20.0, 24.2], [29.8, 35.2], [38.8, 41.1], [45.8, 49.9], [55.8, 57.9], [61.2, 65.6], [74.1, 84.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 17, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001202.mp4", "duration": 45.833, "tgt": [[11.8, 14.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 18, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001319.mp4", "duration": 55.5, "tgt": [[4.2, 10.5], [18.5, 24.8], [28.1, 32.3], [34.4, 38.1], [44.1, 45.7], [49.6, 52.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 19, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001452.mp4", "duration": 72.667, "tgt": [[17.5, 19.4], [26.3, 28.0], [41.3, 43.2], [54.5, 56.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 20, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000173.mp4", "duration": 191.5, "tgt": [[94.7, 98.0], [98.9, 102.3], [118.0, 121.5], [131.3, 134.9], [135.9, 139.2], [178.6, 183.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 21, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000173.mp4", "duration": 191.5, "tgt": [[94.7, 98.0], [98.9, 102.3], [118.0, 121.5], [131.3, 134.9], [135.9, 139.2], [178.6, 183.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 22, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000285.mp4", "duration": 192.833, "tgt": [[53.9, 57.5], [57.8, 62.3], [73.5, 82.5], [82.9, 93.6], [126.8, 134.6], [156.8, 164.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 23, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000285.mp4", "duration": 192.833, "tgt": [[53.9, 57.5], [57.8, 62.3], [73.5, 82.5], [82.9, 93.6], [126.8, 134.6], [156.8, 164.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 24, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000357.mp4", "duration": 126.833, "tgt": [[7.6, 10.0], [19.4, 21.3], [21.6, 24.2], [24.9, 30.2], [43.1, 46.2], [61.9, 64.6], [79.9, 82.5], [86.4, 90.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 25, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000357.mp4", "duration": 126.833, "tgt": [[7.6, 10.0], [19.4, 21.3], [21.6, 24.2], [24.9, 30.2], [43.1, 46.2], [61.9, 64.6], [79.9, 82.5], [86.4, 90.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 26, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000785.mp4", "duration": 87.833, "tgt": [[3.4, 5.6], [43.3, 45.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 27, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000785.mp4", "duration": 87.833, "tgt": [[3.4, 5.6], [43.3, 45.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 28, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000844.mp4", "duration": 169.333, "tgt": [[22.0, 24.6], [94.1, 101.2], [117.6, 121.8], [163.7, 167.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 29, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000844.mp4", "duration": 169.333, "tgt": [[22.0, 24.6], [94.1, 101.2], [117.6, 121.8], [163.7, 167.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 30, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000940.mp4", "duration": 93.167, "tgt": [[35.3, 54.8], [79.4, 83.8], [84.1, 88.8], [89.2, 90.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 31, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000940.mp4", "duration": 93.167, "tgt": [[35.3, 54.8], [68.3, 69.5], [79.4, 83.8], [84.1, 88.8], [89.2, 90.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 32, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001076.mp4", "duration": 64.667, "tgt": [[10.8, 13.4], [14.0, 16.8], [19.0, 21.8], [26.8, 28.7], [29.0, 31.9], [32.6, 37.0], [58.3, 61.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 33, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001076.mp4", "duration": 64.667, "tgt": [[10.8, 13.4], [14.0, 16.8], [19.0, 21.8], [26.8, 28.7], [29.0, 31.9], [32.6, 37.0], [58.3, 61.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 34, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001134.mp4", "duration": 183.833, "tgt": [[5.9, 12.1], [15.4, 20.6], [46.3, 50.9], [66.1, 71.9], [79.4, 82.6], [94.9, 98.2], [101.2, 104.3], [127.5, 145.8], [158.9, 162.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 35, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001134.mp4", "duration": 183.833, "tgt": [[5.9, 12.1], [15.4, 20.6], [46.3, 50.9], [66.1, 71.9], [79.4, 82.6], [94.9, 98.2], [101.2, 104.3], [127.5, 145.8], [158.9, 162.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 36, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001558.mp4", "duration": 142.667, "tgt": [[0.0, 1.9], [45.3, 46.7], [87.2, 88.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 37, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001558.mp4", "duration": 142.667, "tgt": [[0.0, 1.9], [45.3, 46.7], [87.2, 88.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 38, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000011.mp4", "duration": 79.167, "tgt": [[0.4, 12.4], [15.2, 32.0], [40.7, 60.0], [61.8, 77.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 39, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000045.mp4", "duration": 235.333, "tgt": [[32.6, 42.0], [46.5, 62.5], [74.3, 84.8], [126.9, 129.4], [129.6, 130.3], [130.8, 131.6], [131.7, 137.7], [155.6, 165.1], [212.6, 225.1], [225.6, 228.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 40, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000129.mp4", "duration": 191.833, "tgt": [[3.4, 9.1], [11.0, 16.6], [24.4, 29.0], [33.5, 49.4], [59.9, 70.4], [78.3, 94.9], [113.6, 129.0], [141.0, 145.4], [146.2, 148.5], [173.1, 178.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 41, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000367.mp4", "duration": 222.333, "tgt": [[0.0, 3.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'javelin throw'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 42, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000367.mp4", "duration": 222.333, "tgt": [[8.7, 10.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'high jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 43, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000593.mp4", "duration": 56.0, "tgt": [[45.8, 49.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 44, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000691.mp4", "duration": 177.0, "tgt": [[84.7, 88.4], [99.4, 105.7], [117.6, 123.3], [156.9, 160.2], [160.3, 161.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 45, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000765.mp4", "duration": 117.167, "tgt": [[10.1, 24.0], [26.7, 30.1], [34.9, 42.7], [55.9, 70.0], [72.3, 80.7], [87.9, 93.7], [102.9, 113.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 46, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000846.mp4", "duration": 27.167, "tgt": [[1.8, 5.6], [6.4, 9.3], [10.4, 26.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 47, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001066.mp4", "duration": 152.333, "tgt": [[5.3, 10.4], [22.6, 35.7], [57.0, 68.3], [81.9, 85.2], [92.5, 100.4], [132.3, 140.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 48, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001114.mp4", "duration": 48.5, "tgt": [[0.4, 8.4], [20.8, 25.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 49, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001129.mp4", "duration": 67.667, "tgt": [[19.8, 32.3], [41.2, 43.6], [44.1, 45.6], [50.5, 53.2], [54.3, 56.8], [61.9, 64.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 50, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001268.mp4", "duration": 94.667, "tgt": [[7.4, 10.4], [10.8, 12.6], [44.4, 50.7], [58.4, 60.2], [62.7, 64.0], [67.0, 68.0], [74.5, 76.9], [81.2, 87.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 51, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001527.mp4", "duration": 212.667, "tgt": [[8.0, 15.2], [25.5, 28.3], [71.9, 80.3], [94.7, 97.6], [141.3, 147.4], [159.9, 165.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 52, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000058.mp4", "duration": 114.167, "tgt": [[1.9, 26.6], [27.3, 113.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 53, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000504.mp4", "duration": 72.667, "tgt": [[0.3, 66.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 54, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000635.mp4", "duration": 31.5, "tgt": [[0.0, 8.4], [8.8, 17.5], [18.1, 27.0], [27.3, 31.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 55, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000698.mp4", "duration": 17.0, "tgt": [[1.1, 14.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 56, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000737.mp4", "duration": 147.667, "tgt": [[20.1, 43.2], [80.3, 90.9], [116.6, 134.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 57, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000740.mp4", "duration": 411.333, "tgt": [[0.0, 5.6], [5.8, 12.5], [12.5, 18.4], [35.3, 40.9], [43.7, 49.5], [52.1, 56.9], [128.5, 133.8], [133.6, 140.1], [140.1, 144.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 58, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001307.mp4", "duration": 204.667, "tgt": [[12.1, 28.4], [39.0, 54.5], [77.7, 91.6], [103.7, 124.8], [140.3, 160.0], [175.9, 198.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 59, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001468.mp4", "duration": 184.5, "tgt": [[137.2, 143.5], [144.4, 153.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 60, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001483.mp4", "duration": 41.667, "tgt": [[6.1, 13.9], [18.1, 30.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 61, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000374.mp4", "duration": 121.667, "tgt": [[3.1, 4.2], [6.7, 8.3], [89.0, 90.4], [93.2, 96.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'tennis swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 62, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000756.mp4", "duration": 36.0, "tgt": [[10.4, 11.9], [12.3, 13.1], [13.2, 14.0], [14.4, 16.1], [19.1, 19.4], [19.8, 20.2], [22.9, 23.4], [30.8, 31.5], [31.7, 32.2], [34.2, 35.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'tennis swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 63, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001168.mp4", "duration": 109.667, "tgt": [[79.7, 81.1], [98.4, 99.3], [103.6, 104.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'tennis swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 64, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000293.mp4", "duration": 233.5, "tgt": [[195.6, 198.4], [210.2, 216.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'javelin throw'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 65, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000864.mp4", "duration": 195.5, "tgt": [[31.2, 39.5], [40.3, 46.4], [61.9, 69.7], [124.2, 131.2], [132.5, 139.6], [161.1, 164.5], [167.2, 174.5], [187.7, 194.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'javelin throw'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 66, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001209.mp4", "duration": 351.0, "tgt": [[61.0, 66.3], [77.7, 81.4], [156.6, 164.2], [187.5, 197.5], [222.2, 225.2], [250.9, 258.6], [267.7, 273.2], [283.3, 286.9], [311.6, 316.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'javelin throw'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 67, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000444.mp4", "duration": 157.0, "tgt": [[12.4, 22.9], [36.2, 42.7], [55.0, 63.0], [77.0, 85.7], [94.4, 105.3], [113.1, 118.0], [129.6, 138.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'pole vault'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 68, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001162.mp4", "duration": 120.667, "tgt": [[42.7, 64.8], [65.7, 69.4], [70.6, 77.5], [79.3, 82.5], [83.5, 87.4], [89.4, 93.8], [95.1, 99.3], [99.7, 101.7], [102.5, 105.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'pole vault'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 69, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000051.mp4", "duration": 116.0, "tgt": [[51.7, 55.8], [67.3, 70.0], [79.8, 82.4], [93.9, 97.7], [107.4, 110.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'billiards'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 70, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000412.mp4", "duration": 207.333, "tgt": [[165.9, 167.6], [169.5, 173.0], [175.2, 177.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'billiards'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 71, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000556.mp4", "duration": 67.667, "tgt": [[4.8, 5.7], [21.2, 22.5], [33.6, 35.0], [45.9, 50.4], [62.4, 63.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'billiards'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 72, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000577.mp4", "duration": 164.333, "tgt": [[106.9, 111.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'billiards'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 73, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001247.mp4", "duration": 93.167, "tgt": [[85.8, 92.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'billiards'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 74, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001267.mp4", "duration": 126.0, "tgt": [[16.3, 21.6], [29.1, 34.2], [56.4, 59.8], [66.6, 69.3], [118.2, 122.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'billiards'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 75, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000073.mp4", "duration": 98.833, "tgt": [[3.8, 10.1], [21.7, 23.4], [36.6, 43.4], [50.4, 53.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'throw discus'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 76, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000319.mp4", "duration": 138.833, "tgt": [[14.2, 18.2], [47.6, 54.3], [97.7, 104.2], [123.6, 130.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'throw discus'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 77, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000671.mp4", "duration": 116.667, "tgt": [[9.4, 15.1], [16.8, 20.5], [24.4, 31.8], [32.2, 40.5], [41.7, 46.9], [50.2, 56.6], [57.6, 63.5], [63.9, 71.2], [71.9, 80.2], [80.7, 89.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'throw discus'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 78, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000987.mp4", "duration": 184.833, "tgt": [[33.5, 39.1], [45.1, 49.4], [96.6, 102.9], [107.9, 115.3], [128.1, 132.2], [161.8, 168.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'throw discus'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 79, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001080.mp4", "duration": 69.167, "tgt": [[0.0, 23.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'throw discus'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 80, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001223.mp4", "duration": 116.0, "tgt": [[27.4, 33.9], [74.2, 82.7], [102.0, 108.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'throw discus'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 81, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000085.mp4", "duration": 220.667, "tgt": [[170.6, 172.6], [212.7, 217.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 82, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000624.mp4", "duration": 119.833, "tgt": [[0.0, 1.8], [15.8, 19.0], [19.1, 23.1], [27.4, 29.8], [35.7, 37.6], [49.8, 53.9], [58.8, 62.0], [69.7, 74.3], [74.6, 77.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 83, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001081.mp4", "duration": 117.167, "tgt": [[16.2, 18.5], [52.9, 55.8], [105.9, 108.6], [127.5, 130.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 84, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001309.mp4", "duration": 84.833, "tgt": [[2.1, 4.2], [6.6, 13.9], [20.8, 23.7], [29.7, 41.6], [44.0, 52.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 85, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001532.mp4", "duration": 238.833, "tgt": [[4.7, 14.6], [24.2, 28.5], [32.1, 37.5], [38.7, 42.4], [42.5, 45.5], [121.4, 127.1], [213.3, 217.0], [220.4, 221.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 86, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000046.mp4", "duration": 177.833, "tgt": [[79.2, 84.0], [128.1, 132.7], [147.4, 150.9], [170.5, 173.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'golf swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 87, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000113.mp4", "duration": 86.167, "tgt": [[60.8, 79.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'golf swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 88, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000238.mp4", "duration": 102.667, "tgt": [[9.6, 15.7], [21.0, 31.8], [38.5, 78.2], [81.6, 89.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'golf swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 89, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000611.mp4", "duration": 178.667, "tgt": [[88.6, 90.9], [124.1, 129.6], [158.2, 159.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'golf swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 90, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000847.mp4", "duration": 236.0, "tgt": [[0.5, 3.3], [221.6, 232.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'golf swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 91, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001127.mp4", "duration": 180.833, "tgt": [[25.8, 49.8], [50.1, 69.9], [165.6, 169.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'golf swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 92, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001163.mp4", "duration": 100.833, "tgt": [[5.2, 8.0], [22.4, 25.6], [33.8, 40.5], [46.9, 66.0], [78.4, 83.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'golf swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 93, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000188.mp4", "duration": 538.667, "tgt": [[14.0, 17.7], [22.3, 27.2], [91.2, 103.9], [504.6, 510.6], [513.2, 517.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'high jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 94, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000442.mp4", "duration": 155.0, "tgt": [[7.3, 10.9], [11.4, 14.6], [15.7, 20.9], [21.7, 35.3], [50.7, 52.9], [55.3, 57.4], [58.4, 72.0], [134.0, 140.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'high jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 95, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000896.mp4", "duration": 226.667, "tgt": [[18.6, 24.8], [29.1, 34.8], [41.0, 47.4], [57.2, 64.0], [69.4, 78.7], [82.4, 88.6], [89.8, 94.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'high jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 96, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001008.mp4", "duration": 99.5, "tgt": [[7.3, 13.8], [17.5, 20.3], [26.0, 31.6], [37.5, 42.9], [46.1, 49.1], [55.5, 62.5], [68.0, 73.5], [80.0, 86.8], [90.8, 93.3], [95.7, 99.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'high jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 97, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000062.mp4", "duration": 15.0, "tgt": [[1.4, 5.1], [10.3, 14.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'long jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 98, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000242.mp4", "duration": 47.0, "tgt": [[0.0, 6.7], [11.8, 20.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'long jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 99, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000355.mp4", "duration": 646.0, "tgt": [[233.1, 240.1], [245.5, 248.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'long jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 100, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001079.mp4", "duration": 490.833, "tgt": [[350.6, 352.6], [353.1, 355.6], [356.2, 371.4], [372.6, 374.6], [382.1, 386.7], [394.1, 399.3], [459.5, 465.2], [465.3, 479.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'long jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 101, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001219.mp4", "duration": 153.333, "tgt": [[0.4, 5.9], [18.1, 26.1], [140.8, 146.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'long jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 102, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001281.mp4", "duration": 70.5, "tgt": [[36.2, 40.2], [40.8, 52.3], [53.5, 59.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'long jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 103, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000560.mp4", "duration": 50.0, "tgt": [[8.9, 9.7], [20.0, 20.8], [35.0, 36.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'basketball dunk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 104, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000718.mp4", "duration": 40.667, "tgt": [[1.0, 4.5], [29.7, 33.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'basketball dunk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 105, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001343.mp4", "duration": 416.167, "tgt": [[12.0, 13.9], [27.8, 30.4], [49.1, 51.3], [67.8, 70.5], [76.4, 80.7], [138.5, 140.6], [163.3, 166.5], [169.3, 171.9], [302.0, 306.2], [308.4, 311.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'basketball dunk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 106, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000413.mp4", "duration": 42.667, "tgt": [[2.5, 4.8], [6.3, 9.0], [9.6, 11.4], [12.2, 14.5], [17.2, 19.8], [22.2, 27.1], [32.0, 35.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'frisbee catch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 107, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000672.mp4", "duration": 41.0, "tgt": [[2.2, 5.6], [7.7, 10.6], [19.4, 24.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'frisbee catch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 108, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001135.mp4", "duration": 47.667, "tgt": [[7.9, 10.4], [16.7, 27.5], [29.7, 38.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'frisbee catch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 109, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001433.mp4", "duration": 19.667, "tgt": [[4.3, 14.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'frisbee catch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 110, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001512.mp4", "duration": 52.833, "tgt": [[10.7, 17.2], [25.5, 33.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'frisbee catch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 111, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001549.mp4", "duration": 114.833, "tgt": [[3.1, 4.7], [4.9, 5.9], [6.1, 6.8], [7.8, 10.2], [57.7, 61.0], [70.6, 74.5], [107.3, 111.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'frisbee catch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 112, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000292.mp4", "duration": 76.0, "tgt": [[65.9, 67.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 113, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000541.mp4", "duration": 29.333, "tgt": [[1.7, 4.8], [10.1, 12.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 114, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000591.mp4", "duration": 33.5, "tgt": [[10.5, 11.6], [24.3, 25.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 115, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000689.mp4", "duration": 130.0, "tgt": [[62.8, 65.4], [101.6, 103.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 116, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000701.mp4", "duration": 64.667, "tgt": [[0.9, 2.4], [16.2, 26.2], [33.9, 35.8], [41.4, 44.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 117, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001118.mp4", "duration": 70.333, "tgt": [[22.6, 27.5], [48.7, 56.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 118, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001153.mp4", "duration": 128.333, "tgt": [[69.5, 72.1], [97.5, 100.5], [101.9, 106.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 119, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001409.mp4", "duration": 172.667, "tgt": [[19.6, 22.2], [57.7, 59.8], [80.3, 84.6], [100.4, 103.4], [121.1, 123.6], [141.1, 146.7], [163.5, 167.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 120, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001556.mp4", "duration": 103.167, "tgt": [[53.0, 55.3], [78.2, 83.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 121, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000004.mp4", "duration": 33.667, "tgt": [[0.2, 1.1], [11.4, 12.2], [18.6, 20.8], [28.3, 29.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 122, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000004.mp4", "duration": 33.667, "tgt": [[1.0, 1.5], [20.8, 22.3], [30.3, 31.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 123, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000353.mp4", "duration": 72.167, "tgt": [[34.2, 34.8], [35.5, 36.5], [38.2, 39.4], [40.4, 41.8], [47.4, 48.1], [48.8, 49.6], [51.4, 52.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 124, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000392.mp4", "duration": 110.333, "tgt": [[19.1, 21.0], [27.5, 29.2], [48.1, 50.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 125, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000549.mp4", "duration": 71.667, "tgt": [[10.5, 11.9], [14.4, 15.9], [21.9, 23.0], [43.1, 54.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 126, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000549.mp4", "duration": 71.667, "tgt": [[11.4, 13.2], [17.7, 21.0], [22.8, 24.2], [31.7, 33.3], [54.8, 57.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 127, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000569.mp4", "duration": 88.167, "tgt": [[2.2, 3.3], [13.5, 14.4], [48.4, 49.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 128, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000569.mp4", "duration": 88.167, "tgt": [[2.7, 3.8], [49.3, 50.3], [65.0, 66.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 129, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000786.mp4", "duration": 98.0, "tgt": [[3.8, 5.9], [15.9, 20.3], [31.9, 34.1], [52.6, 54.7], [79.9, 82.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 130, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000786.mp4", "duration": 98.0, "tgt": [[33.3, 34.8], [53.5, 54.9], [81.8, 83.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 131, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000796.mp4", "duration": 127.333, "tgt": [[23.3, 24.4], [31.0, 31.9], [37.3, 38.2], [47.3, 48.3], [54.5, 55.6], [61.1, 62.1], [73.0, 74.3], [77.9, 78.8], [89.4, 90.7], [105.1, 106.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 132, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000798.mp4", "duration": 108.667, "tgt": [[87.4, 88.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 133, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000798.mp4", "duration": 108.667, "tgt": [[16.2, 19.8], [35.1, 38.9], [55.8, 59.3], [88.2, 89.6], [92.5, 100.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 134, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000897.mp4", "duration": 120.167, "tgt": [[10.4, 13.4], [17.6, 21.2], [33.3, 37.0], [52.8, 56.5], [65.0, 67.6], [88.6, 91.0], [97.6, 103.3], [103.4, 105.1], [105.2, 107.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 135, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000981.mp4", "duration": 28.0, "tgt": [[4.0, 5.0], [22.4, 25.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 136, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001313.mp4", "duration": 174.0, "tgt": [[21.1, 22.6], [36.1, 37.8], [162.2, 164.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 137, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001358.mp4", "duration": 219.5, "tgt": [[193.1, 194.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 138, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001358.mp4", "duration": 219.5, "tgt": [[0.1, 0.9], [1.2, 1.8], [2.1, 2.8], [129.5, 132.7], [160.8, 164.0], [166.3, 168.3], [180.8, 182.2], [186.0, 191.8], [194.0, 195.0], [199.6, 202.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 139, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001460.mp4", "duration": 29.0, "tgt": [[2.2, 3.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 140, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001460.mp4", "duration": 29.0, "tgt": [[2.9, 3.6], [14.2, 15.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 141, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000443.mp4", "duration": 291.333, "tgt": [[0.0, 7.1], [23.6, 33.4], [48.2, 55.8], [72.9, 79.8], [112.9, 119.8], [133.4, 139.8], [164.2, 171.9], [195.0, 199.9], [237.2, 245.4], [262.6, 270.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'hammer throw'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 142, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001072.mp4", "duration": 284.833, "tgt": [[6.8, 14.7], [38.2, 44.6], [62.4, 76.8], [93.3, 100.1], [116.9, 131.6], [154.3, 160.0], [181.4, 194.4], [247.3, 255.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'hammer throw'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 143, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000448.mp4", "duration": 100.5, "tgt": [[16.0, 17.1], [59.0, 59.8], [62.9, 63.4], [66.1, 67.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 144, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001276.mp4", "duration": 62.333, "tgt": [[8.6, 10.4], [13.0, 15.3], [17.3, 19.3], [21.5, 23.4], [26.4, 27.8], [33.6, 35.5], [41.2, 43.0], [47.0, 48.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}, {"version": 1.0, "idx": 145, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001325.mp4", "duration": 175.167, "tgt": [[36.0, 38.0], [47.8, 51.6], [92.4, 94.5], [150.4, 152.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'. You must represent start and end times in seconds. For example: 'The action happens in 4.2 - 6.8, 7.5 - 10.3, 15.1 - 18.6, and 23.4 - 27.5 seconds'."}]
annotations/txt/tal_thumos15.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/txt/tem_perception_test.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/txt/tem_qvhighlights.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/txt/tvg_charades_sta.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/txt/tvg_qvhighlights.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/txt/vhd_qvhighlights.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/txt/vhd_youtube_highlights.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/vid/dvc_hirest.json ADDED
@@ -0,0 +1 @@
 
 
1
+ [{"version": 1.0, "idx": 0, "task": "dvc", "source": "hirest", "video": "hirest/5V3dI2zp1xA_50_121.mp4", "duration": 71.0, "tgt": [[0, 8], [8, 17], [17, 30], [30, 54], [54, 71]], "g": ["take grinned oats flour", "add carrot and chopped spinach", "add coriander and green chilly", "add salt and water mix well", "heat pan and add fruit salt"], "q": "You are given a video about 'make oatmeal pancake mix'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 1, "task": "dvc", "source": "hirest", "video": "hirest/TcB42a05yzg_33_156.mp4", "duration": 123.0, "tgt": [[0, 11], [11, 18], [18, 19], [19, 29], [29, 37], [37, 51], [51, 100], [100, 123]], "g": ["take thee plate with three candlesticks", "color all with color spray", "add glue to candle sstick", "paste the plate on candle stick", "repeat until the three layer stand completed", "place cup cakes on each plate", "change the candlestick size to our need", "decorate the top cupcake with toy"], "q": "You are given a video about 'make a cupcake stand'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 2, "task": "dvc", "source": "hirest", "video": "hirest/2YZHJOLuA8o_63_216.mp4", "duration": 153.0, "tgt": [[0, 11], [11, 41], [41, 60], [60, 73], [73, 85], [85, 96], [96, 107], [107, 113], [113, 133], [133, 153]], "g": ["take seasonal pan", "spray it and put leafs onit", "spread leafs on pan", "cover empty places on leaf", "add spray on top", "add little bit salt on salt", "add pepper on leaf", "put pan on oven", "cook it for six minutes", "transfer cooked leafs on bowl"], "q": "You are given a video about 'make easy and healthy chips'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 3, "task": "dvc", "source": "hirest", "video": "hirest/Gb-uDDzFRLE_17_49.mp4", "duration": 32.0, "tgt": [[0, 15], [15, 23], [23, 32]], "g": ["take bottom and measure size", "cut straight and well", "cut as slide way"], "q": "You are given a video about 'make giant foam candy canes'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 4, "task": "dvc", "source": "hirest", "video": "hirest/Sb9XLGPJ1yk_25_92.mp4", "duration": 67.0, "tgt": [[0, 11], [11, 21], [21, 37], [37, 43], [43, 51], [51, 56], [56, 63], [63, 67]], "g": ["take dishwashing soap", "take olive oil", "take plate", "pour soap on plate", "pour olive oil on plate", "mix with brush", "lather on hand", "squeeze out makeup"], "q": "You are given a video about 'make your own makeup brush cleanser'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 5, "task": "dvc", "source": "hirest", "video": "hirest/FcRrnEpE4x8_165_439.mp4", "duration": 274.0, "tgt": [[0, 3], [3, 40], [40, 72], [72, 92], [92, 136], [136, 163], [163, 173], [173, 186], [186, 217], [217, 267], [267, 274]], "g": ["zero out the container", "put ink", "mix it around", "spray on the palette", "pull on ink", "flash the garment", "do over print", "do quick flash", "put over print", "cure it", "do a stretch test"], "q": "You are given a video about 'do screen printing on fabric'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 6, "task": "dvc", "source": "hirest", "video": "hirest/-DiQNHSyUfs_34_108.mp4", "duration": 74.0, "tgt": [[0, 11], [11, 18], [18, 38], [38, 58], [58, 74]], "g": ["take some water in plastic bag", "mix vinegar with water", "shut the bag but corner open", "take baking soda in tissue paper", "slip tissue inbag and seal it"], "q": "You are given a video about 'make a plastic bag explode'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 7, "task": "dvc", "source": "hirest", "video": "hirest/gqEbqMKoh9I_33_104.mp4", "duration": 71.0, "tgt": [[0, 18], [18, 37], [37, 56], [56, 67], [67, 71]], "g": ["take paper and put soda onit", "fold paper with soda", "put vinegar with sandwich bag", "put vinegar paper on bag", "shack it well"], "q": "You are given a video about 'make a plastic bag explode'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 8, "task": "dvc", "source": "hirest", "video": "hirest/S6N5RpvChWs_32_113.mp4", "duration": 81.0, "tgt": [[0, 24], [24, 62], [62, 71], [71, 72], [72, 81]], "g": ["take water with vinegar in plastic bags", "take baking soda in tissue paper", "fold paper well", "put it into the bag", "keep away to explode"], "q": "You are given a video about 'make a plastic bag explode'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 9, "task": "dvc", "source": "hirest", "video": "hirest/hHgiALvGm5Y_34_158.mp4", "duration": 124.0, "tgt": [[0, 18], [18, 30], [30, 49], [49, 60], [60, 85], [85, 108], [108, 124]], "g": ["take wanted materials", "put baking soda on paper", "fold soda on bundle", "put soda packet on liquid", "put pocket on ground", "it out on bubbles", "repeat same process on another bag"], "q": "You are given a video about 'make a plastic bag explode'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 10, "task": "dvc", "source": "hirest", "video": "hirest/LKSvFmXLQA0_27_46.mp4", "duration": 19.0, "tgt": [[0, 10], [10, 19]], "g": ["make the fire to use cook", "place large logs into bottom"], "q": "You are given a video about 'use a fire pit'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 11, "task": "dvc", "source": "hirest", "video": "hirest/5OVxioAu6IY_46_163.mp4", "duration": 117.0, "tgt": [[0, 19], [19, 44], [44, 117]], "g": ["arrange the hardware parts", "take the air pipe", "take the nuts and bolds"], "q": "You are given a video about 'use a fire pit'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 12, "task": "dvc", "source": "hirest", "video": "hirest/aHDGMOQgHUE_35_99.mp4", "duration": 64.0, "tgt": [[0, 28], [28, 64]], "g": ["make a fire with coal", "check the gas"], "q": "You are given a video about 'use a fire pit'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 13, "task": "dvc", "source": "hirest", "video": "hirest/Tu9foJWCkgs_47_82.mp4", "duration": 35.0, "tgt": [[0, 10], [10, 16], [16, 25], [25, 35]], "g": ["remove lid", "put handle", "pick grill to put stones", "put the lid back"], "q": "You are given a video about 'use a fire pit'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 14, "task": "dvc", "source": "hirest", "video": "hirest/PtbGXfb6B1I_46_209.mp4", "duration": 163.0, "tgt": [[0, 115], [115, 126], [126, 140], [140, 163]], "g": ["dig the soil", "put the stick inside big hole", "adjust the stick by small hole", "fire the stick"], "q": "You are given a video about 'use a fire pit'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 15, "task": "dvc", "source": "hirest", "video": "hirest/wfiAtfED4K8_41_159.mp4", "duration": 118.0, "tgt": [[0, 8], [8, 13], [13, 24], [24, 30], [30, 39], [39, 59], [59, 118]], "g": ["put the logs into the firepit", "spread the larger logs", "put logs crisscross on top", "put the next size", "put tinder bundle on them", "add some smaller brunches", "start the fire"], "q": "You are given a video about 'use a fire pit'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 16, "task": "dvc", "source": "hirest", "video": "hirest/4zQJzWvRJRA_93_285.mp4", "duration": 192.0, "tgt": [[0, 4], [4, 8], [8, 192]], "g": ["put fence post", "add concrete footing", "keep it above soil"], "q": "You are given a video about 'install a privacy fence'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 17, "task": "dvc", "source": "hirest", "video": "hirest/mqJs3zrKACU_75_246.mp4", "duration": 171.0, "tgt": [[0, 7], [7, 14], [14, 17], [17, 18], [18, 22], [22, 32], [32, 38], [38, 40], [40, 43], [43, 49], [49, 57], [57, 59], [59, 68], [68, 72], [72, 79], [79, 91], [91, 93], [93, 109], [109, 125], [125, 143], [143, 159], [159, 171]], "g": ["heat oil in pan", "add mustard seeds", "add garlic and ginger", "add green chili", "add curry leaves", "saute them", "add turmeric powder", "add beets", "add salt", "saute them again", "cover it to cook", "put fenugreek seeds in pan", "roast it", "add mustard seeds", "roast it again", "crush them", "add crushed seeds into beets", "add asafetida", "cook it", "add red chili powder", "add white vinegar", "stir it"], "q": "You are given a video about 'make spicy sauteed beets with mustard and lemon'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 18, "task": "dvc", "source": "hirest", "video": "hirest/HjEf_JHzqe8_40_157.mp4", "duration": 117.0, "tgt": [[0, 6], [6, 8], [8, 17], [17, 35], [35, 41], [41, 48], [48, 57], [57, 63], [63, 71], [71, 76], [76, 94], [94, 108], [108, 117]], "g": ["remove the front grill", "check condenser", "unbolt the top", "unbolt the lines", "put paper tower in lines", "remove the condenser", "put pag oil", "put new condenser", "put ac oil", "put bolt back", "tighten it up", "remove air out of system", "fill in the refrigerant"], "q": "You are given a video about 'fix your car's air conditioner'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 19, "task": "dvc", "source": "hirest", "video": "hirest/6nnKM_y_vn8_81_358.mp4", "duration": 277.0, "tgt": [[0, 7], [7, 18], [18, 63], [63, 76], [76, 99], [99, 112], [112, 127], [127, 147], [147, 183], [183, 201], [201, 215], [215, 229], [229, 263], [263, 273], [273, 277]], "g": ["add browned beef", "put cooked green bell pepper", "add cooked onions", "add chopped cilantro", "add garlic powder", "add cumin", "add chili powder", "add hot sauce", "mix it around with beans", "add diced tomatoes", "add tomato sauce", "add sugar", "mix it all", "put instant pot on saute", "put it on keep warm later"], "q": "You are given a video about 'make glazed beans'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 20, "task": "dvc", "source": "hirest", "video": "hirest/38H8OKlNA9g_55_178.mp4", "duration": 123.0, "tgt": [[0, 30], [30, 31], [31, 41], [41, 53], [53, 60], [60, 89], [89, 95], [95, 123]], "g": ["open the box of smoker", "check the filling status of charcoal", "add more wood chips if required", "close the box after filling", "ensure air can get around the wood", "check the fire of smoker", "cook the meat well", "check the temperature between the correct degree"], "q": "You are given a video about 'use a smoker'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 21, "task": "dvc", "source": "hirest", "video": "hirest/8KWG-HxHcbE_126_412.mp4", "duration": 286.0, "tgt": [[0, 13], [13, 18], [18, 29], [29, 31], [31, 47], [47, 56], [56, 67], [67, 80], [80, 98], [98, 126], [126, 130], [130, 141], [141, 159], [159, 173], [173, 182], [182, 208], [208, 226], [226, 244], [244, 273], [273, 286]], "g": ["preheat oven", "add vinegar at papaya slices", "spread at gently", "add vinegar for tomato slices", "put it on tray", "put tray on oven", "add oil on pan", "add vegetable on pan", "add salt and mix well", "cook squash until fork tender", "add mixer on jar", "add chicken broken on mixer", "put mixer on pot add chickenoil", "add chock salt and pepper", "add chocki on pot", "add chicken drop and banana pieces", "add chopped cease", "add herbs on pot", "mix all well", "add doll and mix well"], "q": "You are given a video about 'make chorizo'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 22, "task": "dvc", "source": "hirest", "video": "hirest/5Jn7OuXb0FA_25_83.mp4", "duration": 58.0, "tgt": [[0, 2], [2, 10], [10, 23], [23, 35], [35, 51], [51, 58]], "g": ["take your felt", "draw your stencil", "cut it out", "lay it out first", "glue the gems", "glue the pin"], "q": "You are given a video about 'make a brooch'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 23, "task": "dvc", "source": "hirest", "video": "hirest/NULTB33w3Bw_56_282.mp4", "duration": 226.0, "tgt": [[0, 5], [5, 11], [11, 24], [24, 71], [71, 95], [95, 107], [107, 114], [114, 144], [144, 161], [161, 193], [193, 226]], "g": ["put acid on glass", "mix until dissolved set aside", "add glue into mixing bowl:", "add foaming soap into bowl", "add shaving cream into bowl", "mix everything together", "add corn starch into mixture", "add pumps of lotion", "add food coloring into bowl", "add borax solution into bowl", "mix and get right consistency"], "q": "You are given a video about 'make fluffy slime'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 24, "task": "dvc", "source": "hirest", "video": "hirest/xcrmJR_RVIw_31_156.mp4", "duration": 125.0, "tgt": [[0, 79], [79, 125]], "g": ["arrange the flowers in the clutch bags", "add a small yellow petals"], "q": "You are given a video about 'dry flowers using a bag'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 25, "task": "dvc", "source": "hirest", "video": "hirest/NPcyRoWVfL4_43_418.mp4", "duration": 375.0, "tgt": [[0, 25], [25, 59], [59, 91], [91, 150], [150, 185], [185, 192], [192, 237], [237, 266], [266, 375]], "g": ["soak fibers in water", "add white vinegar", "add food colors", "mix it", "put in microwave", "take it out", "let it cool", "take out fibers", "wash them"], "q": "You are given a video about 'set vegetable dyes in clothing'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 26, "task": "dvc", "source": "hirest", "video": "hirest/qnN_OLNAHvc_173_230.mp4", "duration": 57.0, "tgt": [[0, 8], [8, 19], [19, 23], [23, 36], [36, 50], [50, 57]], "g": ["assemble the jig", "align it", "set the blade", "check the depth", "spray water on blade while cutting", "keep cords dry"], "q": "You are given a video about 'cut concrete'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 27, "task": "dvc", "source": "hirest", "video": "hirest/qY9CerJY7FU_80_322.mp4", "duration": 242.0, "tgt": [[0, 37], [37, 77], [77, 179], [179, 242], [242, 242]], "g": ["put egg in the bowl", "apply over the face", "put tissue over it", "dry it", "remove the tissue"], "q": "You are given a video about 'make egg strips'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 28, "task": "dvc", "source": "hirest", "video": "hirest/lpyvvneQImU_81_236.mp4", "duration": 155.0, "tgt": [[0, 2], [2, 8], [8, 14], [14, 21], [21, 37], [37, 43], [43, 44], [44, 46], [46, 51], [51, 56], [56, 60], [60, 70], [70, 96], [96, 155]], "g": ["pre-heat a pan", "add oil to the pan", "add white onion", "sautee the onion", "add garlic", "add apple cider vinegar", "add salt", "add pepper", "add butter", "bring mixture to a little boil", "add brown sugar", "add yellow mustard", "reduce heat and simmer until thickened", "drain mustard sauce"], "q": "You are given a video about 'make a simple mustard sauce'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 29, "task": "dvc", "source": "hirest", "video": "hirest/Ti9ibiKSpf8_13_79.mp4", "duration": 66.0, "tgt": [[0, 11], [11, 25], [25, 33], [33, 39], [39, 66]], "g": ["add mustard", "add honey", "add garlic", "add sriracha sauce", "mix it"], "q": "You are given a video about 'make a simple mustard sauce'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 30, "task": "dvc", "source": "hirest", "video": "hirest/cQ0namNxwWA_99_152.mp4", "duration": 53.0, "tgt": [[0, 9], [9, 12], [12, 13], [13, 16], [16, 22], [22, 26], [26, 30], [30, 34], [34, 35], [35, 53]], "g": ["prepare the ingredients", "beat the egg with cinnamon", "add nutmeg with allspice to taste", "spread some almond meal on plate", "place ham on pineapple piece", "dip that to the beaten egg", "coat pineapple with almond meal", "coat ham with almond meal", "place on heated fry pan with oil", "cook until it has browned"], "q": "You are given a video about 'make pineapple fritters'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 31, "task": "dvc", "source": "hirest", "video": "hirest/P2b7fEpIBIg_48_245.mp4", "duration": 197.0, "tgt": [[0, 50], [50, 97], [97, 125], [125, 149], [149, 180], [180, 197]], "g": ["mix the ingredients", "beat them well", "mix dry and wet ingredients", "put them in the try", "decorate with nuts", "put them in oven"], "q": "You are given a video about 'make pineapple fritters'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 32, "task": "dvc", "source": "hirest", "video": "hirest/WcyDbvsftkI_58_261.mp4", "duration": 203.0, "tgt": [[0, 8], [8, 180], [180, 203]], "g": ["cut all pieces at once", "get a pattern", "paste all the pieces through the pattern"], "q": "You are given a video about 'foundation piece a quilt block'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 33, "task": "dvc", "source": "hirest", "video": "hirest/o69yX5dzvDI_70_301.mp4", "duration": 231.0, "tgt": [[0, 65], [65, 134], [134, 231]], "g": ["fix the design paper", "separate the ribbons", "arrange it like a paper design"], "q": "You are given a video about 'foundation piece a quilt block'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 34, "task": "dvc", "source": "hirest", "video": "hirest/2nkCdciXDa8_55_423.mp4", "duration": 368.0, "tgt": [[0, 79], [79, 91], [91, 109], [109, 120], [120, 137], [137, 162], [162, 178], [178, 194], [194, 197], [197, 200], [200, 213], [213, 216], [216, 255], [255, 309], [309, 346], [346, 353], [353, 368]], "g": ["cut the fabric", "put on template", "put paper clips", "make back stitches", "remove paper clips", "whip stitch them together", "do another one", "attach background piece", "remove the template", "fold it in half", "do the seams", "merge them", "sew the line", "attach pebble piece", "sew the edges", "fill the gaps", "make hexagon edges"], "q": "You are given a video about 'foundation piece a quilt block'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 35, "task": "dvc", "source": "hirest", "video": "hirest/5JpT5RcuFUk_115_251.mp4", "duration": 136.0, "tgt": [[0, 2], [2, 8], [8, 20], [20, 24], [24, 28], [28, 34], [34, 59], [59, 66], [66, 79], [79, 86], [86, 136]], "g": ["get plastic glass table", "put light underneath", "flip over the copy", "use a ruler", "use sharpie to draw", "continue to draw lines", "transfer the numbers of patches", "use double mirror", "lay it on inside", "piece the foundation", "sew the foundation sections"], "q": "You are given a video about 'foundation piece a quilt block'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 36, "task": "dvc", "source": "hirest", "video": "hirest/9Bc1mU0LMF8_82_385.mp4", "duration": 303.0, "tgt": [[0, 12], [12, 37], [37, 100], [100, 106], [106, 129], [129, 145], [145, 166], [166, 173], [173, 177], [177, 179], [179, 189], [189, 202], [202, 231], [231, 238], [238, 271], [271, 274], [274, 301], [301, 303]], "g": ["put scrap", "flip it over", "stitch along line", "trim it", "fold it and iron", "attach piece along seam line", "cut extra fabric", "hold over piece", "add the other piece", "trim the seam", "press with iron", "add other pieces", "attach wide black borders to sides", "attach borders on top and bottom", "draw line to stitch", "flip it over", "tear off paper", "press it"], "q": "You are given a video about 'foundation piece a quilt block'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 37, "task": "dvc", "source": "hirest", "video": "hirest/LX_VHvjunFM_4_57.mp4", "duration": 53.0, "tgt": [[0, 17], [17, 51], [51, 53]], "g": ["things to make krispie buns", "put water in pan,mix chocloate", "put the mix in cups"], "q": "You are given a video about 'make rice krispie buns'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 38, "task": "dvc", "source": "hirest", "video": "hirest/opGK2SjlWJU_57_103.mp4", "duration": 46.0, "tgt": [[0, 25], [25, 40], [40, 46]], "g": ["cut the edges of turnip", "cut into small pieces", "cook with vinegar pepper salt"], "q": "You are given a video about 'cook turnips'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 39, "task": "dvc", "source": "hirest", "video": "hirest/9pybtCnEqEQ_28_125.mp4", "duration": 97.0, "tgt": [[0, 3], [3, 26], [26, 65], [65, 82], [82, 92], [92, 97], [97, 97]], "g": ["make grass using bottom of the", "using pencil draw grass", "paint the grass and dry", "paint the flowers with different colors", "put together flowers and grass", "draw leaves", "leave some message in the card"], "q": "You are given a video about 'make a vintage gift tag'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 40, "task": "dvc", "source": "hirest", "video": "hirest/VvpdSid8ruo_74_124.mp4", "duration": 50.0, "tgt": [[0, 7], [7, 17], [17, 27], [27, 30], [30, 34], [34, 50]], "g": ["paste wrapping paper backside to sheet", "cut the outlined deign", "paste the brown paper in middle", "punch the top of tag", "tie lace in the hole", "write message in the brown sheet"], "q": "You are given a video about 'make a vintage gift tag'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 41, "task": "dvc", "source": "hirest", "video": "hirest/UK3-ZpnogDc_103_347.mp4", "duration": 244.0, "tgt": [[0, 9], [9, 24], [24, 28], [28, 37], [37, 61], [61, 74], [74, 165], [165, 193], [193, 216], [216, 244]], "g": ["put paper", "put on craft bond", "stick them", "mark it", "put glitter tape", "make cut outs", "repeat the process", "punch holes", "take threads", "tie threads on tags"], "q": "You are given a video about 'make a vintage gift tag'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 42, "task": "dvc", "source": "hirest", "video": "hirest/pzvKbNsG-T8_226_326.mp4", "duration": 100.0, "tgt": [[0, 21], [21, 38], [38, 58], [58, 74], [74, 100], [100, 100]], "g": ["take paper collection put on top", "pain color on flowers", "put white patches on edges", "put greeting on inside cover", "put message on paper and send", "create calendar on another method"], "q": "You are given a video about 'make a vintage gift tag'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 43, "task": "dvc", "source": "hirest", "video": "hirest/LFufa6Nfc6Y_68_189.mp4", "duration": 121.0, "tgt": [[0, 4], [4, 19], [19, 29], [29, 41], [41, 49], [49, 64], [64, 84], [84, 121]], "g": ["snow look so good", "fourteen desert are there", "get to find a chopping", "set seven topping here", "put a jelly here", "arrange all jelly into show", "all can be yummy", "set all the candy here"], "q": "You are given a video about 'make snow cones'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 44, "task": "dvc", "source": "hirest", "video": "hirest/iKRdPAsOPPk_19_256.mp4", "duration": 237.0, "tgt": [[0, 5], [5, 13], [13, 23], [23, 63], [63, 75], [75, 98], [98, 163], [163, 237]], "g": ["fold sheet into cone shape", "use three different colors to stick", "tape all cones together", "pin plastic cover inside the cone", "put ices in the jar", "grind it well to make snow", "fill snow to cone sheet", "add some colors for topping"], "q": "You are given a video about 'make snow cones'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 45, "task": "dvc", "source": "hirest", "video": "hirest/ZROZTtnz2Jc_69_224.mp4", "duration": 155.0, "tgt": [[0, 6], [6, 11], [11, 47], [47, 69], [69, 72], [72, 74], [74, 89], [89, 100], [100, 106], [106, 108], [108, 128], [128, 131], [131, 150], [150, 155]], "g": ["cut the chocolate into half pieces", "melt white chocolate cookies in oven", "spread it on sheet", "design toys with colorful cookies", "bake it in oven", "cut the pieces with separate design", "take some skumfiduser marshmallow", "take some chocolate cookies for decoration", "transfer the cookies into plastic bag", "mash it well", "dip mash mellow with chocolate", "dip into the mashed cookies", "insert bamboo stick onto that", "pack it like a gift individually"], "q": "You are given a video about 'make a christmas gift for a teacher'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 46, "task": "dvc", "source": "hirest", "video": "hirest/iFMMxvqemww_105_226.mp4", "duration": 121.0, "tgt": [[0, 37], [37, 79], [79, 121]], "g": ["get a bottle of body spray", "tie the body spray with ribbon", "put body spray into paper bag"], "q": "You are given a video about 'make a christmas gift for a teacher'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 47, "task": "dvc", "source": "hirest", "video": "hirest/DBBJe_kdLDI_99_168.mp4", "duration": 69.0, "tgt": [[0, 10], [10, 23], [23, 33], [33, 49], [49, 56], [56, 69]], "g": ["use trim", "put stamps on top", "use gold paper", "make flowers for bin", "use pearls", "use sweet peppermint collection"], "q": "You are given a video about 'make a christmas gift for a teacher'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 48, "task": "dvc", "source": "hirest", "video": "hirest/ZujIOKsOnzw_27_133.mp4", "duration": 106.0, "tgt": [[0, 36], [36, 47], [47, 60], [60, 81], [81, 102], [102, 106]], "g": ["take an materials", "take designed chart and santa", "take tissue and candy on bottle", "take some glitters", "fold glitter on tissue", "put all candy tissue on bottle"], "q": "You are given a video about 'make a christmas gift for a teacher'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 49, "task": "dvc", "source": "hirest", "video": "hirest/gWR8Nsf_Z_g_61_221.mp4", "duration": 160.0, "tgt": [[0, 18], [18, 41], [41, 62], [62, 86], [86, 88], [88, 100], [100, 114], [114, 136], [136, 146], [146, 160]], "g": ["unscrew the top", "lift the top off", "empty the container", "use scoop bottle", "get resin", "put in container", "add water", "put filter back", "screw the top tight", "connect the taps"], "q": "You are given a video about 'clean resin'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 50, "task": "dvc", "source": "hirest", "video": "hirest/yH3p_oMRyns_144_498.mp4", "duration": 354.0, "tgt": [[0, 20], [20, 39], [39, 82], [82, 122], [122, 129], [129, 228], [228, 263], [263, 305], [305, 324], [324, 354]], "g": ["attach dahlia and garden roses", "add spray roses", "add greens", "add brunia", "put another dahlia", "add another rose", "put floral tape", "wrap the bouquet", "cut the ribbon", "put pins"], "q": "You are given a video about 'make a bouquet'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 51, "task": "dvc", "source": "hirest", "video": "hirest/oDRoXgINJAU_108_288.mp4", "duration": 180.0, "tgt": [[0, 13], [13, 25], [25, 50], [50, 70], [70, 76], [76, 89], [89, 99], [99, 120], [120, 140], [140, 148], [148, 168], [168, 180]], "g": ["boil water and add meat on", "fry vegetables using pan", "add carrot on pan", "take out meat", "add sous on boiled meat water", "add cabbage into water", "add carrot on it", "mix all well", "add sous on it", "add tomato and needed herbs", "cook it well", "serve on another bowl"], "q": "You are given a video about 'make borsh russian soup'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 52, "task": "dvc", "source": "hirest", "video": "hirest/sk9KApzKeLw_54_120.mp4", "duration": 66.0, "tgt": [[0, 6], [6, 12], [12, 14], [14, 22], [22, 66]], "g": ["hold the coin between the tips of your fingers", "reach your right hand over the coin to conceal it", "drop the coin into your left palm", "pull your right hand away", "open your empty right hand slowly"], "q": "You are given a video about 'do the 'french drop''. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 53, "task": "dvc", "source": "hirest", "video": "hirest/nN6ZOXoBjps_57_130.mp4", "duration": 73.0, "tgt": [[0, 72], [72, 73]], "g": ["rub off the dust and oxide", "dry immediately"], "q": "You are given a video about 'clean chrome'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 54, "task": "dvc", "source": "hirest", "video": "hirest/YY_JX65mWuw_67_154.mp4", "duration": 87.0, "tgt": [[0, 6], [6, 45], [45, 57], [57, 87]], "g": ["pour cleaning liquid in cloth", "scrub chrome with cloth", "take more liquid to clean more stain", "wipe it by dry cloth"], "q": "You are given a video about 'clean chrome'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 55, "task": "dvc", "source": "hirest", "video": "hirest/daclapuOQnE_73_157.mp4", "duration": 84.0, "tgt": [[0, 9], [9, 25], [25, 41], [41, 74], [74, 84]], "g": ["take scrub and spray water", "rub out the scratches", "use cloth to clean out", "spray water and clean with water", "use scrubber to clean it well"], "q": "You are given a video about 'clean chrome'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 56, "task": "dvc", "source": "hirest", "video": "hirest/qE5eodQCqqc_22_119.mp4", "duration": 97.0, "tgt": [[0, 16], [16, 76], [76, 97]], "g": ["take metal polish into cloth", "wiping down chrome gently", "clean by new cloth fully"], "q": "You are given a video about 'clean chrome'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 57, "task": "dvc", "source": "hirest", "video": "hirest/qPzWyc-XTQ4_37_283.mp4", "duration": 246.0, "tgt": [[0, 4], [4, 35], [35, 44], [44, 78], [78, 90], [90, 96], [96, 199], [199, 208], [208, 226], [226, 229], [229, 237], [237, 246]], "g": ["shake and spray it", "keep doing the process", "let it sit", "rub it around", "wash with water", "dry it off", "repeat the process", "polish it", "use sandpaper for roughness", "use water proof sandpaper", "polish it more", "use different sandpapers"], "q": "You are given a video about 'clean chrome'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 58, "task": "dvc", "source": "hirest", "video": "hirest/ymAgcwbWo38_28_47.mp4", "duration": 19.0, "tgt": [[0, 8], [8, 12], [12, 19]], "g": ["apply liquid on parts", "rub it out the dirty", "use cloth to clean out"], "q": "You are given a video about 'clean chrome'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 59, "task": "dvc", "source": "hirest", "video": "hirest/94Kea14_1I0_58_321.mp4", "duration": 263.0, "tgt": [[0, 10], [10, 21], [21, 56], [56, 263]], "g": ["mark and put lines on paper", "fold it properly", "fix it with tape", "add some color papers"], "q": "You are given a video about 'make a duct tape wallet'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 60, "task": "dvc", "source": "hirest", "video": "hirest/PiIfYPwuw7c_6_368.mp4", "duration": 362.0, "tgt": [[0, 4], [4, 11], [11, 27], [27, 33], [33, 42], [42, 61], [61, 75], [75, 85], [85, 104], [104, 122], [122, 146], [146, 154], [154, 204], [204, 219], [219, 230], [230, 233], [233, 286], [286, 350], [350, 362]], "g": ["leave all tape on borders", "trim it", "cut squares out", "fold the divider", "put it on", "cut a slit", "fold it over", "line the other on edge", "fold the tape", "trim the lower part", "fold the tape", "grab coin pouch", "attach the tape", "cut and fold the tape", "do same on other side", "fold the flap down", "fold rest of tape", "put velcro", "put little tape on edges"], "q": "You are given a video about 'make a duct tape wallet'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 61, "task": "dvc", "source": "hirest", "video": "hirest/AEkQeihliZM_29_95.mp4", "duration": 66.0, "tgt": [[0, 15], [15, 35], [35, 58], [58, 66]], "g": ["arrange wanted length on room", "using spray to spread water", "using cloth to clean out floor", "use machine to clean out"], "q": "You are given a video about 'apply diatomaceous earth to carpet'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 62, "task": "dvc", "source": "hirest", "video": "hirest/n_h7GGgUhDM_73_270.mp4", "duration": 197.0, "tgt": [[0, 20], [20, 32], [32, 82], [82, 137], [137, 176], [176, 197]], "g": ["pick the plant", "select a container with drainage", "fill up with water", "put it on sunny side", "put water not too much", "dry out some time"], "q": "You are given a video about 'grow fresh herbs'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 63, "task": "dvc", "source": "hirest", "video": "hirest/yWpSs787IRI_55_220.mp4", "duration": 165.0, "tgt": [[0, 14], [14, 34], [34, 49], [49, 66], [66, 94], [94, 114], [114, 137], [137, 156], [156, 165]], "g": ["use elastic to create belly", "take pins and thread", "put elastic on dress back", "put three inch elastic on side", "stich elastic using threads and pins", "put stich on other side elastic", "put stich on back ward direction", "tight out elastic on sides", "wear dress as comfort"], "q": "You are given a video about 'make a full belly dance skirt'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 64, "task": "dvc", "source": "hirest", "video": "hirest/Dz_BZ48BhsA_194_535.mp4", "duration": 341.0, "tgt": [[0, 7], [7, 11], [11, 19], [19, 43], [43, 50], [50, 92], [92, 119], [119, 145], [145, 148], [148, 177], [177, 181], [181, 185], [185, 226], [226, 341]], "g": ["measure the size", "leave extra on sides", "cut the edging", "do same with other", "make into regular bra", "sew the edging", "put on bra", "sew bra strap inside", "hook it all", "adjust and cut it", "make thin belt first", "hook it with hip scarf", "embellish the scarf", "add rhinestones"], "q": "You are given a video about 'make a full belly dance skirt'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 65, "task": "dvc", "source": "hirest", "video": "hirest/qGCVZTvjMFw_29_259.mp4", "duration": 230.0, "tgt": [[0, 59], [59, 117], [117, 209], [209, 230]], "g": ["cut the material", "cut like mermaid dress", "mark into different pieces", "stich it"], "q": "You are given a video about 'make a full belly dance skirt'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 66, "task": "dvc", "source": "hirest", "video": "hirest/IX86Gre7d2Y_49_129.mp4", "duration": 80.0, "tgt": [[0, 1], [1, 9], [9, 21], [21, 26], [26, 27], [27, 48], [48, 51], [51, 80]], "g": ["turn skirt inside out", "locate the seams", "draw lines", "sew along the lines", "do back for no side seam", "make dart and stitch", "take piece of elastic", "stitch it at back of skirt"], "q": "You are given a video about 'make a full belly dance skirt'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 67, "task": "dvc", "source": "hirest", "video": "hirest/WIHPzX3zLNo_53_99.mp4", "duration": 46.0, "tgt": [[0, 14], [14, 20], [20, 46]], "g": ["peel the duct tape into two", "joint tape one by one", "repeat the step upto make bowl"], "q": "You are given a video about 'make a duct tape bowl'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 68, "task": "dvc", "source": "hirest", "video": "hirest/LH6B1Ye8hYw_17_80.mp4", "duration": 63.0, "tgt": [[0, 13], [13, 14], [14, 18], [18, 50], [50, 53], [53, 63]], "g": ["take one cup of heavy cream", "add two tsp of buttermilk", "wrap it with plastic cover", "keep it room temperature for eight hours", "transfer into other bowl", "keep it in fridge"], "q": "You are given a video about 'make creme fraiche'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 69, "task": "dvc", "source": "hirest", "video": "hirest/8_Mg-mOvvMw_67_191.mp4", "duration": 124.0, "tgt": [[0, 13], [13, 28], [28, 29], [29, 40], [40, 48], [48, 68], [68, 109], [109, 124]], "g": ["cut roma tomatoes", "spray oil , salt and bake", "cut beetroot and garlic", "add tahini and lemon juice", "add canon oil,salt and mash it", "prepare toppings with parsley", "cut the bread into slices and toast", "assemble everything with omelette"], "q": "You are given a video about 'create a brunch menu'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 70, "task": "dvc", "source": "hirest", "video": "hirest/6C0mL9n9Nko_11_116.mp4", "duration": 105.0, "tgt": [[0, 29], [29, 48], [48, 59], [59, 74], [74, 98], [98, 105]], "g": ["have a needed things", "using a clay design", "another clay", "start pour gycerlin", "remove outer layer of clay", "using chain measure the band size"], "q": "You are given a video about 'make a fairy headband'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 71, "task": "dvc", "source": "hirest", "video": "hirest/vxDNQjk75dU_56_168.mp4", "duration": 112.0, "tgt": [[0, 26], [26, 43], [43, 50], [50, 66], [66, 73], [73, 77], [77, 82], [82, 91], [91, 96], [96, 101], [101, 104], [104, 112]], "g": ["put almond flour in bowl", "add confectioners", "add chocolate chips", "add salt", "add peanut butter", "add melted coconut oil", "add vanilla extract", "mix it", "make a cookie dough", "scoop out the cookie dough", "place on baking sheet", "put in fridge"], "q": "You are given a video about 'make peanut butter cup cookie bites'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 72, "task": "dvc", "source": "hirest", "video": "hirest/WqVQSC5zMBw_20_30.mp4", "duration": 10.0, "tgt": [[0, 5], [5, 10]], "g": ["put mashed potato into freezer", "using bag to store as well"], "q": "You are given a video about 'freeze mashed potatoes'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 73, "task": "dvc", "source": "hirest", "video": "hirest/e76KWpbPdfY_107_465.mp4", "duration": 358.0, "tgt": [[0, 36], [36, 70], [70, 116], [116, 136], [136, 165], [165, 174], [174, 180], [180, 196], [196, 220], [220, 227], [227, 229], [229, 240], [240, 255], [255, 258], [258, 280], [280, 315], [315, 321], [321, 325], [325, 337], [337, 351], [351, 358]], "g": ["take large piece", "feel vertical bars", "take small hinge", "feel the top bar", "put small piece over large", "do both same time", "grab corner tightly", "put pin on right", "stretch the large piece", "put bottom bar on hook", "hold it there", "stretch the hinges", "use blade", "get key cap", "push it on sides", "turn the large key", "put wire on bottom hinges", "turn key over", "open up the wire", "put wire under", "turn key over and push"], "q": "You are given a video about 'fix sticky keyboard keys'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}, {"version": 1.0, "idx": 74, "task": "dvc", "source": "hirest", "video": "hirest/L8kJRfPMAEY_35_115.mp4", "duration": 80.0, "tgt": [[0, 1], [1, 3], [3, 8], [8, 12], [12, 14], [14, 36], [36, 45], [45, 47], [47, 80]], "g": ["pour cereal", "add rice krispies and fruity pebbles", "add marshmallows", "mix it around", "put it in the microwave", "watch the heat", "use hot gloves to take out", "put treats in the dish", "mold the treats"], "q": "You are given a video about 'make rice krispies paint brush treats'. Watch the video carefully and densly describe all the steps. For each step, you need to determine the start and ends times and provide a concise description. The format of your response should be: '<start time> - <end time>, <description>'."}]
annotations/vid/dvc_youcook2.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/vid/eca_charades_sta.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/vid/epm_eg4d_nlq.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/vid/evs_summe.json ADDED
@@ -0,0 +1 @@
 
 
1
+ [{"version": 1.0, "idx": 0, "task": "evs", "source": "summe", "video": "summe/saving_dolphines.mp4", "duration": 223.333, "tgt": [[3.37, 5.172], [5.939, 6.907], [7.708, 9.643], [23.023, 26.427], [27.127, 33.601], [36.47, 43.911], [49.684, 52.386], [57.058, 61.295], [65.7, 68.336], [76.677, 80.782], [110.111, 112.981], [134.302, 138.073], [201.637, 202.004]], "q": "You are given a video about saving dolphines. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 1, "task": "evs", "source": "summe", "video": "summe/bearpark_climbing.mp4", "duration": 134.0, "tgt": [[3.4, 4.48], [9.6, 12.6], [48.081, 48.361], [48.641, 50.481], [76.842, 80.322], [83.962, 85.922], [85.962, 87.162], [101.322, 102.442], [103.482, 106.642], [106.682, 107.082], [107.562, 107.602], [109.362, 109.722], [109.962, 110.322], [110.922, 112.003], [112.043, 112.563], [124.443, 125.963]], "q": "You are given a video about bearpark climbing. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 2, "task": "evs", "source": "summe", "video": "summe/playing_ball.mp4", "duration": 104.333, "tgt": [[1.267, 4.267], [5.767, 5.967], [11.267, 11.467], [11.501, 12.701], [12.867, 13.367], [13.867, 22.334], [29.435, 31.202], [53.203, 53.869], [53.903, 57.536], [86.671, 87.704]], "q": "You are given a video about playing ball. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 3, "task": "evs", "source": "summe", "video": "summe/car_railcrossing.mp4", "duration": 169.667, "tgt": [[3.337, 4.271], [10.444, 12.013], [13.347, 13.715], [14.015, 26.194], [78.45, 86.625], [86.725, 87.893], [112.386, 113.153], [157.166, 158.334], [158.635, 161.237]], "q": "You are given a video about car railcrossing. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 4, "task": "evs", "source": "summe", "video": "summe/st_maarten_landing.mp4", "duration": 70.333, "tgt": [[25.32, 28.52], [48.2, 55.88]], "q": "You are given a video about st maarten landing. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 5, "task": "evs", "source": "summe", "video": "summe/excavators_river_crossing.mp4", "duration": 389.167, "tgt": [[2.52, 9.72], [10.08, 12.96], [27.0, 29.36], [35.48, 50.521], [62.641, 63.201], [69.201, 83.441], [93.681, 101.921], [102.801, 114.241], [114.641, 115.041], [119.241, 121.801], [122.401, 123.201], [377.844, 378.284]], "q": "You are given a video about excavators river crossing. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 6, "task": "evs", "source": "summe", "video": "summe/uncut_evening_flight.mp4", "duration": 323.0, "tgt": [[0.0, 15.049], [17.285, 18.453], [65.835, 66.536], [68.972, 69.039], [69.873, 72.075], [88.726, 95.199], [111.016, 117.156], [163.771, 170.445], [286.566, 287.467], [287.567, 288.735], [289.169, 290.737], [307.054, 320.802]], "q": "You are given a video about uncut evening flight. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 7, "task": "evs", "source": "summe", "video": "summe/paluma_jump.mp4", "duration": 86.167, "tgt": [[40.044, 45.083], [59.966, 66.706], [78.619, 79.454], [81.623, 82.457]], "q": "You are given a video about paluma jump. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 8, "task": "evs", "source": "summe", "video": "summe/statue_of_liberty.mp4", "duration": 154.833, "tgt": [[24.174, 35.778], [41.418, 42.587], [43.513, 43.675], [45.971, 47.019], [47.059, 49.597], [57.333, 57.655], [97.422, 100.887], [113.86, 117.204], [117.285, 119.138], [121.757, 121.999], [122.2, 122.684], [132.796, 135.939]], "q": "You are given a video about statue of liberty. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 9, "task": "evs", "source": "summe", "video": "summe/air_force_one.mp4", "duration": 180.167, "tgt": [[56.964, 57.204], [66.804, 67.164], [68.125, 68.925], [70.405, 85.926], [155.77, 171.531], [173.772, 174.412]], "q": "You are given a video about air force one. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 10, "task": "evs", "source": "summe", "video": "summe/eiffel_tower.mp4", "duration": 199.167, "tgt": [[3.095, 6.713], [30.712, 31.516], [47.676, 56.037], [79.433, 81.644], [100.055, 101.1], [101.623, 104.356], [113.723, 114.446], [115.572, 119.592], [126.345, 131.008], [134.305, 138.968], [195.447, 198.623]], "q": "You are given a video about eiffel tower. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 11, "task": "evs", "source": "summe", "video": "summe/kids_playing_in_leaves.mp4", "duration": 106.667, "tgt": [[2.336, 14.683], [15.484, 20.055], [30.967, 32.836], [42.313, 44.482], [53.492, 54.059], [68.175, 68.308], [71.679, 73.914], [91.434, 92.068], [93.035, 94.137], [96.773, 97.741]], "q": "You are given a video about kids playing in leaves. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 12, "task": "evs", "source": "summe", "video": "summe/car_over_camera.mp4", "duration": 146.5, "tgt": [[36.743, 37.377], [38.011, 38.511], [56.766, 59.269], [60.77, 64.808], [65.609, 65.776], [85.799, 92.941], [92.974, 93.742], [101.551, 102.419], [102.852, 107.591], [108.392, 112.03], [140.863, 143.766]], "q": "You are given a video about car over camera. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 13, "task": "evs", "source": "summe", "video": "summe/fire_domino.mp4", "duration": 54.0, "tgt": [[9.132, 13.206], [13.274, 13.375], [18.841, 19.86], [23.865, 24.612], [25.461, 25.733], [44.574, 47.357]], "q": "You are given a video about fire domino. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 14, "task": "evs", "source": "summe", "video": "summe/cockpit_landing.mp4", "duration": 302.167, "tgt": [[5.414, 6.918], [38.298, 41.673], [72.652, 83.179], [111.05, 120.574], [240.112, 244.223], [244.724, 247.364], [276.505, 277.608], [281.05, 284.358], [285.695, 301.034]], "q": "You are given a video about cockpit landing. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 15, "task": "evs", "source": "summe", "video": "summe/valparaiso_downhill.mp4", "duration": 173.167, "tgt": [[10.278, 16.284], [23.525, 24.559], [28.697, 29.932], [31.4, 31.467], [34.136, 35.738], [35.938, 35.971], [37.273, 37.64], [39.475, 42.145], [66.504, 69.14], [69.874, 70.141], [89.028, 92.431], [96.269, 97.67], [127.769, 129.737], [136.044, 137.779], [160.036, 166.743], [166.843, 167.01]], "q": "You are given a video about valparaiso downhill. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 16, "task": "evs", "source": "summe", "video": "summe/jumps.mp4", "duration": 38.333, "tgt": [[2.069, 3.517], [9.062, 9.352], [9.683, 10.635], [12.083, 15.601], [24.456, 24.704]], "q": "You are given a video about jumps. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 17, "task": "evs", "source": "summe", "video": "summe/base_jumping.mp4", "duration": 158.167, "tgt": [[4.232, 6.113], [27.037, 33.049], [42.89, 49.204], [50.38, 50.581], [50.615, 53.503], [68.92, 69.793], [81.817, 86.855], [95.99, 96.192], [96.293, 97.502], [98.61, 100.961], [122.994, 128.939], [135.555, 136.16], [136.429, 136.731]], "q": "You are given a video about base jumping. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 18, "task": "evs", "source": "summe", "video": "summe/bike_polo.mp4", "duration": 102.5, "tgt": [[21.619, 26.09], [28.041, 31.974], [62.2, 62.503], [63.142, 63.444], [70.269, 75.615], [85.702, 88.123]], "q": "You are given a video about bike polo. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 19, "task": "evs", "source": "summe", "video": "summe/notre_dame.mp4", "duration": 192.333, "tgt": [[4.459, 9.668], [11.002, 11.502], [17.586, 17.92], [21.754, 22.337], [23.212, 24.254], [24.629, 25.088], [89.098, 95.308], [96.808, 98.016], [118.728, 118.811], [123.187, 131.23], [137.731, 140.398], [142.357, 142.607], [143.024, 144.524], [176.696, 180.53]], "q": "You are given a video about notre dame. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 20, "task": "evs", "source": "summe", "video": "summe/playing_on_water_slide.mp4", "duration": 102.667, "tgt": [[6.975, 8.51], [10.913, 10.98], [11.08, 11.113], [15.018, 15.285], [28.634, 34.407], [34.474, 34.674], [34.741, 36.543], [45.053, 48.057], [52.128, 52.896], [64.142, 65.41], [91.541, 91.875], [92.042, 99.717]], "q": "You are given a video about playing on water slide. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 21, "task": "evs", "source": "summe", "video": "summe/paintball.mp4", "duration": 254.667, "tgt": [[7.088, 15.183], [36.153, 40.683], [55.278, 63.205], [97.68, 106.32], [230.507, 236.378], [238.224, 245.06]], "q": "You are given a video about paintball. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 22, "task": "evs", "source": "summe", "video": "summe/bus_in_rock_tunnel.mp4", "duration": 171.5, "tgt": [[47.114, 52.619], [53.053, 55.956], [67.667, 69.403], [69.569, 70.537], [75.976, 76.343], [84.184, 84.518], [85.318, 85.385], [85.819, 86.119], [88.054, 97.464], [125.759, 126.226], [139.472, 140.106], [140.173, 140.507], [146.146, 147.914], [149.015, 149.082], [166.332, 168.735]], "q": "You are given a video about bus in rock tunnel. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 23, "task": "evs", "source": "summe", "video": "summe/scuba.mp4", "duration": 74.333, "tgt": [[1.6, 2.333], [6.533, 6.9], [15.7, 19.167], [19.7, 21.567], [36.833, 40.4], [55.266, 57.833], [62.3, 62.633], [63.8, 66.1], [71.933, 72.2], [72.466, 73.866]], "q": "You are given a video about scuba. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 24, "task": "evs", "source": "summe", "video": "summe/cooking.mp4", "duration": 86.167, "tgt": [[30.968, 31.574], [32.92, 39.181], [51.838, 54.194], [68.13, 69.005], [69.072, 74.054]], "q": "You are given a video about cooking. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}]
annotations/vid/evs_tvsum.json ADDED
@@ -0,0 +1 @@
 
 
1
+ [{"version": 1.0, "idx": 0, "task": "evs", "source": "tvsum", "video": "tvsum/akI8YFjEmUw.mp4", "duration": 133.333, "tgt": [[32, 40], [44, 46], [66, 68], [110, 112], [122, 126], [128, 130]], "q": "You are given a video about getting vehicle unstuck. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 1, "task": "evs", "source": "tvsum", "video": "tvsum/J0nA4VgnoCo.mp4", "duration": 584.667, "tgt": [[94, 100], [110, 112], [240, 246], [248, 252], [266, 276], [290, 308], [320, 334], [336, 344], [346, 350], [352, 356], [364, 370], [374, 376], [410, 414]], "q": "You are given a video about changing vehicle tire. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 2, "task": "evs", "source": "tvsum", "video": "tvsum/kLxoNp-UchI.mp4", "duration": 130.0, "tgt": [[8, 14], [26, 28], [50, 52], [84, 86], [96, 98], [100, 102], [106, 108], [118, 120]], "q": "You are given a video about dog show. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 3, "task": "evs", "source": "tvsum", "video": "tvsum/WxtbjNsCQ8A.mp4", "duration": 265.5, "tgt": [[66, 76], [82, 84], [92, 100], [122, 128], [140, 142], [172, 174], [222, 224], [232, 240]], "q": "You are given a video about bee keeping. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 4, "task": "evs", "source": "tvsum", "video": "tvsum/HT5vyqe0Xaw.mp4", "duration": 322.667, "tgt": [[38, 40], [42, 50], [58, 60], [64, 68], [82, 86], [110, 112], [120, 122], [148, 152], [158, 166], [168, 170], [236, 238], [258, 260], [276, 278], [280, 282], [284, 286], [298, 302]], "q": "You are given a video about getting vehicle unstuck. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 5, "task": "evs", "source": "tvsum", "video": "tvsum/gzDbaEs1Rlg.mp4", "duration": 288.333, "tgt": [[34, 36], [70, 72], [76, 78], [104, 110], [114, 116], [120, 122], [136, 140], [144, 150], [154, 164], [172, 176], [204, 206], [222, 224]], "q": "You are given a video about changing vehicle tire. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 6, "task": "evs", "source": "tvsum", "video": "tvsum/EE-bNr36nyA.mp4", "duration": 98.167, "tgt": [[30, 36], [42, 44], [46, 50], [80, 82], [86, 88]], "q": "You are given a video about bee keeping. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 7, "task": "evs", "source": "tvsum", "video": "tvsum/eQu1rNs0an0.mp4", "duration": 164.5, "tgt": [[4, 6], [12, 14], [22, 24], [48, 52], [64, 72], [106, 112], [136, 140]], "q": "You are given a video about attempting bike tricks. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 8, "task": "evs", "source": "tvsum", "video": "tvsum/oDXZc0tZe04.mp4", "duration": 380.833, "tgt": [[2, 4], [92, 96], [98, 100], [112, 118], [152, 154], [156, 158], [200, 210], [250, 256], [294, 300], [302, 312], [334, 340], [364, 366], [368, 370]], "q": "You are given a video about bee keeping. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 9, "task": "evs", "source": "tvsum", "video": "tvsum/Hl-__g2gn_A.mp4", "duration": 243.833, "tgt": [[8, 10], [62, 64], [72, 74], [76, 80], [84, 86], [110, 114], [128, 132], [166, 174], [186, 188], [206, 208], [224, 230]], "q": "You are given a video about making sandwich. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 10, "task": "evs", "source": "tvsum", "video": "tvsum/WG0MBPpPC6I.mp4", "duration": 397.667, "tgt": [[8, 10], [52, 54], [106, 108], [126, 128], [168, 170], [180, 182], [190, 194], [222, 224], [242, 244], [256, 260], [266, 270], [274, 278], [284, 288], [300, 302], [324, 330], [338, 348], [352, 354], [356, 358], [366, 368]], "q": "You are given a video about making sandwich. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 11, "task": "evs", "source": "tvsum", "video": "tvsum/GsAD1KT1xo8.mp4", "duration": 145.333, "tgt": [[76, 80], [86, 90], [92, 96], [114, 118], [122, 124], [128, 130], [136, 138]], "q": "You are given a video about parkour. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 12, "task": "evs", "source": "tvsum", "video": "tvsum/xxdtq8mxegs.mp4", "duration": 144.167, "tgt": [[70, 88], [90, 92], [102, 104]], "q": "You are given a video about grooming an animal. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 13, "task": "evs", "source": "tvsum", "video": "tvsum/NyBmCxDoHJU.mp4", "duration": 189.667, "tgt": [[20, 24], [40, 44], [60, 62], [66, 70], [106, 108], [114, 116], [118, 122], [128, 130], [136, 142], [176, 178]], "q": "You are given a video about dog show. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 14, "task": "evs", "source": "tvsum", "video": "tvsum/Bhxk-O1Y7Ho.mp4", "duration": 450.833, "tgt": [[0, 2], [74, 82], [102, 104], [108, 112], [114, 122], [160, 170], [174, 176], [268, 270], [272, 274], [276, 278], [286, 288], [292, 296], [298, 300], [326, 328], [330, 334], [336, 340], [434, 440], [442, 450]], "q": "You are given a video about grooming an animal. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 15, "task": "evs", "source": "tvsum", "video": "tvsum/_xMr-HKMfVA.mp4", "duration": 148.833, "tgt": [[18, 20], [60, 62], [80, 90], [94, 100], [118, 122]], "q": "You are given a video about flash mob gathering. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 16, "task": "evs", "source": "tvsum", "video": "tvsum/i3wAGJaaktw.mp4", "duration": 156.833, "tgt": [[4, 8], [10, 12], [28, 32], [56, 58], [64, 66], [88, 100]], "q": "You are given a video about grooming an animal. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 17, "task": "evs", "source": "tvsum", "video": "tvsum/JKpqYvAdIsw.mp4", "duration": 152.0, "tgt": [[18, 20], [40, 42], [48, 50], [52, 60], [108, 110], [124, 126], [128, 132], [140, 142]], "q": "You are given a video about flash mob gathering. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 18, "task": "evs", "source": "tvsum", "video": "tvsum/byxOvuiIJV0.mp4", "duration": 154.5, "tgt": [[30, 34], [36, 40], [48, 50], [66, 68], [78, 80], [98, 100], [102, 108], [128, 132]], "q": "You are given a video about flash mob gathering. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 19, "task": "evs", "source": "tvsum", "video": "tvsum/RBCABdttQmI.mp4", "duration": 364.333, "tgt": [[6, 10], [32, 36], [84, 86], [88, 90], [124, 126], [216, 218], [222, 236], [266, 268], [272, 276], [310, 318], [326, 328], [336, 342], [346, 348], [352, 354], [356, 358]], "q": "You are given a video about parade. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 20, "task": "evs", "source": "tvsum", "video": "tvsum/XzYM3PfTM4w.mp4", "duration": 111.0, "tgt": [[20, 24], [32, 34], [48, 58], [68, 70]], "q": "You are given a video about changing vehicle tire. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 21, "task": "evs", "source": "tvsum", "video": "tvsum/0tmA_C6XwfM.mp4", "duration": 141.333, "tgt": [[28, 30], [52, 78], [106, 108]], "q": "You are given a video about grooming an animal. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 22, "task": "evs", "source": "tvsum", "video": "tvsum/qqR6AEXwxoQ.mp4", "duration": 269.333, "tgt": [[56, 62], [80, 86], [118, 120], [148, 164], [166, 174], [240, 246]], "q": "You are given a video about attempting bike tricks. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 23, "task": "evs", "source": "tvsum", "video": "tvsum/vdmoEJ5YbrQ.mp4", "duration": 329.333, "tgt": [[32, 34], [64, 66], [84, 86], [96, 98], [100, 112], [126, 128], [132, 138], [140, 156], [178, 180], [184, 188]], "q": "You are given a video about getting vehicle unstuck. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 24, "task": "evs", "source": "tvsum", "video": "tvsum/LRw_obCPUt0.mp4", "duration": 260.333, "tgt": [[4, 8], [58, 68], [118, 132], [136, 138], [156, 160], [234, 242]], "q": "You are given a video about making sandwich. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 25, "task": "evs", "source": "tvsum", "video": "tvsum/uGu_10sucQo.mp4", "duration": 167.0, "tgt": [[30, 32], [36, 38], [40, 42], [64, 66], [72, 74], [76, 78], [96, 108], [112, 120], [124, 126]], "q": "You are given a video about bee keeping. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 26, "task": "evs", "source": "tvsum", "video": "tvsum/4wU_LUjG5Ic.mp4", "duration": 167.0, "tgt": [[8, 14], [16, 22], [52, 56], [58, 60], [102, 108], [138, 140]], "q": "You are given a video about parade. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 27, "task": "evs", "source": "tvsum", "video": "tvsum/98MoyGZKHXc.mp4", "duration": 187.5, "tgt": [[28, 30], [64, 70], [76, 86], [102, 108], [116, 118], [128, 130]], "q": "You are given a video about changing vehicle tire. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 28, "task": "evs", "source": "tvsum", "video": "tvsum/XkqCExn6_Us.mp4", "duration": 187.833, "tgt": [[2, 4], [22, 24], [26, 28], [56, 58], [64, 66], [74, 84], [86, 88], [92, 94], [100, 102], [106, 108]], "q": "You are given a video about parkour. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 29, "task": "evs", "source": "tvsum", "video": "tvsum/Se3oxnaPsz0.mp4", "duration": 138.833, "tgt": [[42, 44], [52, 56], [76, 78], [86, 90], [102, 106], [110, 116]], "q": "You are given a video about bee keeping. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 30, "task": "evs", "source": "tvsum", "video": "tvsum/sTEELN-vY30.mp4", "duration": 149.0, "tgt": [[0, 14], [66, 74], [118, 122], [130, 132]], "q": "You are given a video about getting vehicle unstuck. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 31, "task": "evs", "source": "tvsum", "video": "tvsum/JgHubY5Vw3Y.mp4", "duration": 143.667, "tgt": [[12, 14], [80, 82], [94, 112]], "q": "You are given a video about attempting bike tricks. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 32, "task": "evs", "source": "tvsum", "video": "tvsum/3eYKfiOEJNs.mp4", "duration": 194.167, "tgt": [[50, 58], [60, 62], [64, 66], [76, 78], [82, 92], [170, 174], [176, 178]], "q": "You are given a video about grooming an animal. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 33, "task": "evs", "source": "tvsum", "video": "tvsum/xwqBXPGE9pQ.mp4", "duration": 233.833, "tgt": [[20, 22], [30, 34], [58, 62], [70, 74], [76, 88], [106, 108], [122, 130], [166, 172], [174, 176], [224, 226]], "q": "You are given a video about getting vehicle unstuck. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 34, "task": "evs", "source": "tvsum", "video": "tvsum/VuWGsYPqAX8.mp4", "duration": 216.5, "tgt": [[62, 64], [84, 88], [92, 94], [112, 116], [118, 120], [122, 126], [130, 132], [142, 144], [178, 180], [182, 184], [194, 198], [204, 206], [212, 214]], "q": "You are given a video about flash mob gathering. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 35, "task": "evs", "source": "tvsum", "video": "tvsum/jcoYJXDG9sw.mp4", "duration": 199.167, "tgt": [[90, 94], [96, 108], [122, 126], [158, 168]], "q": "You are given a video about dog show. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 36, "task": "evs", "source": "tvsum", "video": "tvsum/z_6gVvQb2d0.mp4", "duration": 276.333, "tgt": [[132, 136], [138, 144], [152, 154], [160, 162], [168, 170], [172, 174], [190, 192], [196, 198], [202, 204], [212, 220], [224, 228], [232, 234], [236, 238], [240, 242], [246, 248]], "q": "You are given a video about parade. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 37, "task": "evs", "source": "tvsum", "video": "tvsum/-esJrBWj2d8.mp4", "duration": 230.333, "tgt": [[18, 20], [42, 54], [56, 58], [64, 80], [110, 114], [116, 118], [126, 128]], "q": "You are given a video about dog show. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 38, "task": "evs", "source": "tvsum", "video": "tvsum/xmEERLqJ2kU.mp4", "duration": 446.0, "tgt": [[230, 232], [234, 236], [240, 244], [248, 250], [252, 258], [260, 262], [266, 268], [270, 272], [278, 280], [326, 340], [356, 358], [374, 394], [396, 398], [400, 404], [410, 414]], "q": "You are given a video about flash mob gathering. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 39, "task": "evs", "source": "tvsum", "video": "tvsum/b626MiF1ew4.mp4", "duration": 235.833, "tgt": [[0, 2], [18, 20], [48, 50], [58, 60], [62, 64], [72, 76], [80, 82], [112, 114], [124, 126], [128, 130], [152, 154], [156, 158], [192, 194], [208, 210], [212, 214], [218, 222]], "q": "You are given a video about parkour. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 40, "task": "evs", "source": "tvsum", "video": "tvsum/EYqVtI9YWJA.mp4", "duration": 198.167, "tgt": [[98, 100], [114, 116], [138, 140], [144, 146], [160, 174], [176, 182], [188, 190]], "q": "You are given a video about attempting bike tricks. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 41, "task": "evs", "source": "tvsum", "video": "tvsum/iVt07TCkFM0.mp4", "duration": 104.333, "tgt": [[8, 10], [44, 46], [52, 56], [58, 60], [66, 70], [72, 74]], "q": "You are given a video about attempting bike tricks. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 42, "task": "evs", "source": "tvsum", "video": "tvsum/fWutDQy1nnY.mp4", "duration": 585.833, "tgt": [[90, 92], [150, 154], [156, 158], [188, 194], [196, 216], [224, 228], [236, 238], [242, 244], [254, 256], [270, 274], [276, 280], [282, 290], [296, 298], [338, 340], [344, 346], [348, 354], [356, 360], [364, 366], [368, 374], [382, 384], [386, 388], [412, 416]], "q": "You are given a video about parade. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 43, "task": "evs", "source": "tvsum", "video": "tvsum/37rzWOQsNIw.mp4", "duration": 191.667, "tgt": [[0, 6], [16, 22], [62, 64], [66, 72], [74, 82], [154, 158], [162, 164]], "q": "You are given a video about making sandwich. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 44, "task": "evs", "source": "tvsum", "video": "tvsum/91IHQYk1IQM.mp4", "duration": 110.5, "tgt": [[18, 20], [38, 40], [42, 46], [52, 54], [62, 66], [82, 84], [86, 88]], "q": "You are given a video about parade. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 45, "task": "evs", "source": "tvsum", "video": "tvsum/cjibtmSLxQ4.mp4", "duration": 647.5, "tgt": [[8, 10], [30, 32], [38, 46], [60, 64], [74, 76], [78, 84], [122, 126], [146, 148], [150, 152], [172, 180], [182, 184], [190, 192], [198, 202], [258, 264], [272, 276], [322, 324], [380, 382], [384, 390], [518, 522], [524, 526], [528, 532], [534, 538], [578, 582], [584, 586], [588, 594], [606, 608], [618, 622], [628, 630]], "q": "You are given a video about parkour. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 46, "task": "evs", "source": "tvsum", "video": "tvsum/PJrm840pAUI.mp4", "duration": 274.167, "tgt": [[2, 4], [28, 30], [32, 34], [42, 48], [114, 116], [132, 134], [166, 168], [174, 184], [192, 194], [196, 202], [246, 248], [250, 252], [254, 262]], "q": "You are given a video about parkour. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 47, "task": "evs", "source": "tvsum", "video": "tvsum/Yi4Ij2NM7U4.mp4", "duration": 405.5, "tgt": [[58, 62], [124, 126], [130, 132], [134, 140], [142, 144], [156, 162], [166, 176], [236, 238], [240, 242], [254, 270], [294, 304], [306, 308], [334, 336], [370, 378]], "q": "You are given a video about making sandwich. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 48, "task": "evs", "source": "tvsum", "video": "tvsum/AwmHb44_ouw.mp4", "duration": 353.667, "tgt": [[0, 2], [94, 98], [102, 108], [120, 126], [138, 140], [142, 144], [152, 154], [156, 160], [164, 166], [172, 174], [188, 190], [192, 194], [198, 202], [208, 210], [222, 224], [256, 258], [260, 262], [274, 276], [284, 286], [300, 302]], "q": "You are given a video about changing vehicle tire. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 49, "task": "evs", "source": "tvsum", "video": "tvsum/E11zDS9XGzg.mp4", "duration": 510.167, "tgt": [[40, 44], [90, 94], [110, 114], [130, 134], [158, 166], [168, 172], [174, 176], [178, 180], [242, 252], [260, 266], [268, 272], [288, 292], [298, 300], [304, 308], [318, 324], [378, 380], [388, 392], [422, 424], [440, 450], [456, 458]], "q": "You are given a video about dog show. Watch the video carefully and summarize it into multiple short segments. The total length of the segments should be about 15% of the original video. The format of your response should be: 'The summary locates in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}]
annotations/vid/gvq_qa_ego4d.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/vid/rar_perception_test.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/vid/rvq_star.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/vid/slc_cross_task.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/vid/slc_ht_step.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/vid/tal_perception_test.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/vid/tal_thumos14.json ADDED
@@ -0,0 +1 @@
 
 
1
+ [{"version": 1.0, "idx": 0, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000324.mp4", "duration": 149.0, "tgt": [[49.2, 53.5], [116.7, 122.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 1, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000664.mp4", "duration": 86.0, "tgt": [[1.3, 4.8], [22.3, 25.3], [56.8, 59.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 2, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000714.mp4", "duration": 179.5, "tgt": [[26.2, 29.9], [37.0, 39.1], [141.4, 144.7], [150.1, 152.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 3, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000767.mp4", "duration": 77.167, "tgt": [[10.9, 12.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 4, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000873.mp4", "duration": 34.833, "tgt": [[6.3, 9.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 5, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000946.mp4", "duration": 27.833, "tgt": [[0.9, 3.5], [13.1, 17.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 6, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000964.mp4", "duration": 63.333, "tgt": [[7.5, 15.3], [42.5, 47.3], [52.2, 56.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 7, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001038.mp4", "duration": 121.167, "tgt": [[50.6, 54.2], [110.5, 114.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 8, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001182.mp4", "duration": 119.333, "tgt": [[1.3, 3.7], [78.1, 81.2], [89.9, 92.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 9, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001447.mp4", "duration": 229.833, "tgt": [[172.7, 175.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'baseball pitch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 10, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000006.mp4", "duration": 67.0, "tgt": [[18.8, 57.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 11, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000211.mp4", "duration": 98.833, "tgt": [[8.2, 10.1], [10.8, 12.5], [13.1, 14.8], [17.3, 19.5], [30.5, 40.1], [43.0, 44.2], [47.6, 49.3], [50.8, 52.5], [55.3, 62.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 12, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000426.mp4", "duration": 153.0, "tgt": [[27.1, 29.4], [34.9, 35.9], [53.1, 54.1], [55.6, 58.1], [63.2, 64.6], [78.9, 80.3], [84.3, 86.3], [100.1, 101.3], [144.1, 145.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 13, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000450.mp4", "duration": 28.333, "tgt": [[10.9, 13.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 14, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000615.mp4", "duration": 206.5, "tgt": [[10.0, 16.6], [52.6, 54.6], [111.9, 119.0], [128.4, 134.8], [171.4, 173.9], [184.0, 188.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 15, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000762.mp4", "duration": 133.0, "tgt": [[3.0, 4.8], [13.3, 15.8], [26.8, 29.1], [47.9, 50.1], [56.5, 58.0], [74.4, 76.8], [83.6, 85.8], [98.3, 101.6], [114.2, 118.4], [129.7, 132.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 16, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001078.mp4", "duration": 97.5, "tgt": [[0.5, 2.2], [4.6, 9.0], [15.3, 17.0], [20.0, 24.2], [29.8, 35.2], [38.8, 41.1], [45.8, 49.9], [55.8, 57.9], [61.2, 65.6], [74.1, 84.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 17, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001202.mp4", "duration": 45.833, "tgt": [[11.8, 14.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 18, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001319.mp4", "duration": 55.5, "tgt": [[4.2, 10.5], [18.5, 24.8], [28.1, 32.3], [34.4, 38.1], [44.1, 45.7], [49.6, 52.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 19, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001452.mp4", "duration": 72.667, "tgt": [[17.5, 19.4], [26.3, 28.0], [41.3, 43.2], [54.5, 56.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'volleyball spiking'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 20, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000173.mp4", "duration": 191.5, "tgt": [[94.7, 98.0], [98.9, 102.3], [118.0, 121.5], [131.3, 134.9], [135.9, 139.2], [178.6, 183.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 21, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000173.mp4", "duration": 191.5, "tgt": [[94.7, 98.0], [98.9, 102.3], [118.0, 121.5], [131.3, 134.9], [135.9, 139.2], [178.6, 183.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 22, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000285.mp4", "duration": 192.833, "tgt": [[53.9, 57.5], [57.8, 62.3], [73.5, 82.5], [82.9, 93.6], [126.8, 134.6], [156.8, 164.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 23, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000285.mp4", "duration": 192.833, "tgt": [[53.9, 57.5], [57.8, 62.3], [73.5, 82.5], [82.9, 93.6], [126.8, 134.6], [156.8, 164.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 24, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000357.mp4", "duration": 126.833, "tgt": [[7.6, 10.0], [19.4, 21.3], [21.6, 24.2], [24.9, 30.2], [43.1, 46.2], [61.9, 64.6], [79.9, 82.5], [86.4, 90.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 25, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000357.mp4", "duration": 126.833, "tgt": [[7.6, 10.0], [19.4, 21.3], [21.6, 24.2], [24.9, 30.2], [43.1, 46.2], [61.9, 64.6], [79.9, 82.5], [86.4, 90.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 26, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000785.mp4", "duration": 87.833, "tgt": [[3.4, 5.6], [43.3, 45.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 27, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000785.mp4", "duration": 87.833, "tgt": [[3.4, 5.6], [43.3, 45.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 28, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000844.mp4", "duration": 169.333, "tgt": [[22.0, 24.6], [94.1, 101.2], [117.6, 121.8], [163.7, 167.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 29, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000844.mp4", "duration": 169.333, "tgt": [[22.0, 24.6], [94.1, 101.2], [117.6, 121.8], [163.7, 167.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 30, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000940.mp4", "duration": 93.167, "tgt": [[35.3, 54.8], [79.4, 83.8], [84.1, 88.8], [89.2, 90.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 31, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000940.mp4", "duration": 93.167, "tgt": [[35.3, 54.8], [68.3, 69.5], [79.4, 83.8], [84.1, 88.8], [89.2, 90.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 32, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001076.mp4", "duration": 64.667, "tgt": [[10.8, 13.4], [14.0, 16.8], [19.0, 21.8], [26.8, 28.7], [29.0, 31.9], [32.6, 37.0], [58.3, 61.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 33, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001076.mp4", "duration": 64.667, "tgt": [[10.8, 13.4], [14.0, 16.8], [19.0, 21.8], [26.8, 28.7], [29.0, 31.9], [32.6, 37.0], [58.3, 61.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 34, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001134.mp4", "duration": 183.833, "tgt": [[5.9, 12.1], [15.4, 20.6], [46.3, 50.9], [66.1, 71.9], [79.4, 82.6], [94.9, 98.2], [101.2, 104.3], [127.5, 145.8], [158.9, 162.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 35, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001134.mp4", "duration": 183.833, "tgt": [[5.9, 12.1], [15.4, 20.6], [46.3, 50.9], [66.1, 71.9], [79.4, 82.6], [94.9, 98.2], [101.2, 104.3], [127.5, 145.8], [158.9, 162.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 36, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001558.mp4", "duration": 142.667, "tgt": [[0.0, 1.9], [45.3, 46.7], [87.2, 88.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cliff diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 37, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001558.mp4", "duration": 142.667, "tgt": [[0.0, 1.9], [45.3, 46.7], [87.2, 88.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 38, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000011.mp4", "duration": 79.167, "tgt": [[0.4, 12.4], [15.2, 32.0], [40.7, 60.0], [61.8, 77.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 39, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000045.mp4", "duration": 235.333, "tgt": [[32.6, 42.0], [46.5, 62.5], [74.3, 84.8], [126.9, 129.4], [129.6, 130.3], [130.8, 131.6], [131.7, 137.7], [155.6, 165.1], [212.6, 225.1], [225.6, 228.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 40, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000129.mp4", "duration": 191.833, "tgt": [[3.4, 9.1], [11.0, 16.6], [24.4, 29.0], [33.5, 49.4], [59.9, 70.4], [78.3, 94.9], [113.6, 129.0], [141.0, 145.4], [146.2, 148.5], [173.1, 178.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 41, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000367.mp4", "duration": 222.333, "tgt": [[0.0, 3.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'javelin throw'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 42, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000367.mp4", "duration": 222.333, "tgt": [[8.7, 10.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'high jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 43, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000593.mp4", "duration": 56.0, "tgt": [[45.8, 49.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 44, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000691.mp4", "duration": 177.0, "tgt": [[84.7, 88.4], [99.4, 105.7], [117.6, 123.3], [156.9, 160.2], [160.3, 161.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 45, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000765.mp4", "duration": 117.167, "tgt": [[10.1, 24.0], [26.7, 30.1], [34.9, 42.7], [55.9, 70.0], [72.3, 80.7], [87.9, 93.7], [102.9, 113.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 46, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000846.mp4", "duration": 27.167, "tgt": [[1.8, 5.6], [6.4, 9.3], [10.4, 26.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 47, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001066.mp4", "duration": 152.333, "tgt": [[5.3, 10.4], [22.6, 35.7], [57.0, 68.3], [81.9, 85.2], [92.5, 100.4], [132.3, 140.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 48, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001114.mp4", "duration": 48.5, "tgt": [[0.4, 8.4], [20.8, 25.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 49, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001129.mp4", "duration": 67.667, "tgt": [[19.8, 32.3], [41.2, 43.6], [44.1, 45.6], [50.5, 53.2], [54.3, 56.8], [61.9, 64.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 50, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001268.mp4", "duration": 94.667, "tgt": [[7.4, 10.4], [10.8, 12.6], [44.4, 50.7], [58.4, 60.2], [62.7, 64.0], [67.0, 68.0], [74.5, 76.9], [81.2, 87.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 51, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001527.mp4", "duration": 212.667, "tgt": [[8.0, 15.2], [25.5, 28.3], [71.9, 80.3], [94.7, 97.6], [141.3, 147.4], [159.9, 165.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'shotput'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 52, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000058.mp4", "duration": 114.167, "tgt": [[1.9, 26.6], [27.3, 113.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 53, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000504.mp4", "duration": 72.667, "tgt": [[0.3, 66.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 54, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000635.mp4", "duration": 31.5, "tgt": [[0.0, 8.4], [8.8, 17.5], [18.1, 27.0], [27.3, 31.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 55, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000698.mp4", "duration": 17.0, "tgt": [[1.1, 14.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 56, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000737.mp4", "duration": 147.667, "tgt": [[20.1, 43.2], [80.3, 90.9], [116.6, 134.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 57, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000740.mp4", "duration": 411.333, "tgt": [[0.0, 5.6], [5.8, 12.5], [12.5, 18.4], [35.3, 40.9], [43.7, 49.5], [52.1, 56.9], [128.5, 133.8], [133.6, 140.1], [140.1, 144.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 58, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001307.mp4", "duration": 204.667, "tgt": [[12.1, 28.4], [39.0, 54.5], [77.7, 91.6], [103.7, 124.8], [140.3, 160.0], [175.9, 198.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 59, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001468.mp4", "duration": 184.5, "tgt": [[137.2, 143.5], [144.4, 153.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 60, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001483.mp4", "duration": 41.667, "tgt": [[6.1, 13.9], [18.1, 30.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'clean and jerk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 61, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000374.mp4", "duration": 121.667, "tgt": [[3.1, 4.2], [6.7, 8.3], [89.0, 90.4], [93.2, 96.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'tennis swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 62, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000756.mp4", "duration": 36.0, "tgt": [[10.4, 11.9], [12.3, 13.1], [13.2, 14.0], [14.4, 16.1], [19.1, 19.4], [19.8, 20.2], [22.9, 23.4], [30.8, 31.5], [31.7, 32.2], [34.2, 35.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'tennis swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 63, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001168.mp4", "duration": 109.667, "tgt": [[79.7, 81.1], [98.4, 99.3], [103.6, 104.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'tennis swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 64, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000293.mp4", "duration": 233.5, "tgt": [[195.6, 198.4], [210.2, 216.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'javelin throw'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 65, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000864.mp4", "duration": 195.5, "tgt": [[31.2, 39.5], [40.3, 46.4], [61.9, 69.7], [124.2, 131.2], [132.5, 139.6], [161.1, 164.5], [167.2, 174.5], [187.7, 194.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'javelin throw'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 66, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001209.mp4", "duration": 351.0, "tgt": [[61.0, 66.3], [77.7, 81.4], [156.6, 164.2], [187.5, 197.5], [222.2, 225.2], [250.9, 258.6], [267.7, 273.2], [283.3, 286.9], [311.6, 316.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'javelin throw'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 67, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000444.mp4", "duration": 157.0, "tgt": [[12.4, 22.9], [36.2, 42.7], [55.0, 63.0], [77.0, 85.7], [94.4, 105.3], [113.1, 118.0], [129.6, 138.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'pole vault'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 68, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001162.mp4", "duration": 120.667, "tgt": [[42.7, 64.8], [65.7, 69.4], [70.6, 77.5], [79.3, 82.5], [83.5, 87.4], [89.4, 93.8], [95.1, 99.3], [99.7, 101.7], [102.5, 105.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'pole vault'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 69, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000051.mp4", "duration": 116.0, "tgt": [[51.7, 55.8], [67.3, 70.0], [79.8, 82.4], [93.9, 97.7], [107.4, 110.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'billiards'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 70, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000412.mp4", "duration": 207.333, "tgt": [[165.9, 167.6], [169.5, 173.0], [175.2, 177.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'billiards'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 71, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000556.mp4", "duration": 67.667, "tgt": [[4.8, 5.7], [21.2, 22.5], [33.6, 35.0], [45.9, 50.4], [62.4, 63.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'billiards'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 72, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000577.mp4", "duration": 164.333, "tgt": [[106.9, 111.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'billiards'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 73, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001247.mp4", "duration": 93.167, "tgt": [[85.8, 92.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'billiards'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 74, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001267.mp4", "duration": 126.0, "tgt": [[16.3, 21.6], [29.1, 34.2], [56.4, 59.8], [66.6, 69.3], [118.2, 122.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'billiards'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 75, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000073.mp4", "duration": 98.833, "tgt": [[3.8, 10.1], [21.7, 23.4], [36.6, 43.4], [50.4, 53.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'throw discus'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 76, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000319.mp4", "duration": 138.833, "tgt": [[14.2, 18.2], [47.6, 54.3], [97.7, 104.2], [123.6, 130.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'throw discus'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 77, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000671.mp4", "duration": 116.667, "tgt": [[9.4, 15.1], [16.8, 20.5], [24.4, 31.8], [32.2, 40.5], [41.7, 46.9], [50.2, 56.6], [57.6, 63.5], [63.9, 71.2], [71.9, 80.2], [80.7, 89.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'throw discus'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 78, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000987.mp4", "duration": 184.833, "tgt": [[33.5, 39.1], [45.1, 49.4], [96.6, 102.9], [107.9, 115.3], [128.1, 132.2], [161.8, 168.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'throw discus'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 79, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001080.mp4", "duration": 69.167, "tgt": [[0.0, 23.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'throw discus'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 80, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001223.mp4", "duration": 116.0, "tgt": [[27.4, 33.9], [74.2, 82.7], [102.0, 108.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'throw discus'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 81, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000085.mp4", "duration": 220.667, "tgt": [[170.6, 172.6], [212.7, 217.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 82, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000624.mp4", "duration": 119.833, "tgt": [[0.0, 1.8], [15.8, 19.0], [19.1, 23.1], [27.4, 29.8], [35.7, 37.6], [49.8, 53.9], [58.8, 62.0], [69.7, 74.3], [74.6, 77.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 83, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001081.mp4", "duration": 117.167, "tgt": [[16.2, 18.5], [52.9, 55.8], [105.9, 108.6], [127.5, 130.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 84, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001309.mp4", "duration": 84.833, "tgt": [[2.1, 4.2], [6.6, 13.9], [20.8, 23.7], [29.7, 41.6], [44.0, 52.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 85, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001532.mp4", "duration": 238.833, "tgt": [[4.7, 14.6], [24.2, 28.5], [32.1, 37.5], [38.7, 42.4], [42.5, 45.5], [121.4, 127.1], [213.3, 217.0], [220.4, 221.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'diving'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 86, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000046.mp4", "duration": 177.833, "tgt": [[79.2, 84.0], [128.1, 132.7], [147.4, 150.9], [170.5, 173.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'golf swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 87, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000113.mp4", "duration": 86.167, "tgt": [[60.8, 79.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'golf swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 88, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000238.mp4", "duration": 102.667, "tgt": [[9.6, 15.7], [21.0, 31.8], [38.5, 78.2], [81.6, 89.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'golf swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 89, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000611.mp4", "duration": 178.667, "tgt": [[88.6, 90.9], [124.1, 129.6], [158.2, 159.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'golf swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 90, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000847.mp4", "duration": 236.0, "tgt": [[0.5, 3.3], [221.6, 232.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'golf swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 91, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001127.mp4", "duration": 180.833, "tgt": [[25.8, 49.8], [50.1, 69.9], [165.6, 169.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'golf swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 92, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001163.mp4", "duration": 100.833, "tgt": [[5.2, 8.0], [22.4, 25.6], [33.8, 40.5], [46.9, 66.0], [78.4, 83.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'golf swing'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 93, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000188.mp4", "duration": 538.667, "tgt": [[14.0, 17.7], [22.3, 27.2], [91.2, 103.9], [504.6, 510.6], [513.2, 517.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'high jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 94, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000442.mp4", "duration": 155.0, "tgt": [[7.3, 10.9], [11.4, 14.6], [15.7, 20.9], [21.7, 35.3], [50.7, 52.9], [55.3, 57.4], [58.4, 72.0], [134.0, 140.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'high jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 95, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000896.mp4", "duration": 226.667, "tgt": [[18.6, 24.8], [29.1, 34.8], [41.0, 47.4], [57.2, 64.0], [69.4, 78.7], [82.4, 88.6], [89.8, 94.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'high jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 96, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001008.mp4", "duration": 99.5, "tgt": [[7.3, 13.8], [17.5, 20.3], [26.0, 31.6], [37.5, 42.9], [46.1, 49.1], [55.5, 62.5], [68.0, 73.5], [80.0, 86.8], [90.8, 93.3], [95.7, 99.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'high jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 97, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000062.mp4", "duration": 15.0, "tgt": [[1.4, 5.1], [10.3, 14.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'long jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 98, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000242.mp4", "duration": 47.0, "tgt": [[0.0, 6.7], [11.8, 20.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'long jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 99, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000355.mp4", "duration": 646.0, "tgt": [[233.1, 240.1], [245.5, 248.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'long jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 100, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001079.mp4", "duration": 490.833, "tgt": [[350.6, 352.6], [353.1, 355.6], [356.2, 371.4], [372.6, 374.6], [382.1, 386.7], [394.1, 399.3], [459.5, 465.2], [465.3, 479.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'long jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 101, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001219.mp4", "duration": 153.333, "tgt": [[0.4, 5.9], [18.1, 26.1], [140.8, 146.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'long jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 102, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001281.mp4", "duration": 70.5, "tgt": [[36.2, 40.2], [40.8, 52.3], [53.5, 59.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'long jump'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 103, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000560.mp4", "duration": 50.0, "tgt": [[8.9, 9.7], [20.0, 20.8], [35.0, 36.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'basketball dunk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 104, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000718.mp4", "duration": 40.667, "tgt": [[1.0, 4.5], [29.7, 33.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'basketball dunk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 105, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001343.mp4", "duration": 416.167, "tgt": [[12.0, 13.9], [27.8, 30.4], [49.1, 51.3], [67.8, 70.5], [76.4, 80.7], [138.5, 140.6], [163.3, 166.5], [169.3, 171.9], [302.0, 306.2], [308.4, 311.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'basketball dunk'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 106, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000413.mp4", "duration": 42.667, "tgt": [[2.5, 4.8], [6.3, 9.0], [9.6, 11.4], [12.2, 14.5], [17.2, 19.8], [22.2, 27.1], [32.0, 35.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'frisbee catch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 107, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000672.mp4", "duration": 41.0, "tgt": [[2.2, 5.6], [7.7, 10.6], [19.4, 24.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'frisbee catch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 108, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001135.mp4", "duration": 47.667, "tgt": [[7.9, 10.4], [16.7, 27.5], [29.7, 38.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'frisbee catch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 109, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001433.mp4", "duration": 19.667, "tgt": [[4.3, 14.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'frisbee catch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 110, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001512.mp4", "duration": 52.833, "tgt": [[10.7, 17.2], [25.5, 33.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'frisbee catch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 111, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001549.mp4", "duration": 114.833, "tgt": [[3.1, 4.7], [4.9, 5.9], [6.1, 6.8], [7.8, 10.2], [57.7, 61.0], [70.6, 74.5], [107.3, 111.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'frisbee catch'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 112, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000292.mp4", "duration": 76.0, "tgt": [[65.9, 67.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 113, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000541.mp4", "duration": 29.333, "tgt": [[1.7, 4.8], [10.1, 12.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 114, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000591.mp4", "duration": 33.5, "tgt": [[10.5, 11.6], [24.3, 25.9]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 115, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000689.mp4", "duration": 130.0, "tgt": [[62.8, 65.4], [101.6, 103.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 116, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000701.mp4", "duration": 64.667, "tgt": [[0.9, 2.4], [16.2, 26.2], [33.9, 35.8], [41.4, 44.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 117, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001118.mp4", "duration": 70.333, "tgt": [[22.6, 27.5], [48.7, 56.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 118, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001153.mp4", "duration": 128.333, "tgt": [[69.5, 72.1], [97.5, 100.5], [101.9, 106.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 119, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001409.mp4", "duration": 172.667, "tgt": [[19.6, 22.2], [57.7, 59.8], [80.3, 84.6], [100.4, 103.4], [121.1, 123.6], [141.1, 146.7], [163.5, 167.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 120, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001556.mp4", "duration": 103.167, "tgt": [[53.0, 55.3], [78.2, 83.8]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'soccer penalty'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 121, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000004.mp4", "duration": 33.667, "tgt": [[0.2, 1.1], [11.4, 12.2], [18.6, 20.8], [28.3, 29.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 122, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000004.mp4", "duration": 33.667, "tgt": [[1.0, 1.5], [20.8, 22.3], [30.3, 31.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 123, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000353.mp4", "duration": 72.167, "tgt": [[34.2, 34.8], [35.5, 36.5], [38.2, 39.4], [40.4, 41.8], [47.4, 48.1], [48.8, 49.6], [51.4, 52.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 124, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000392.mp4", "duration": 110.333, "tgt": [[19.1, 21.0], [27.5, 29.2], [48.1, 50.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 125, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000549.mp4", "duration": 71.667, "tgt": [[10.5, 11.9], [14.4, 15.9], [21.9, 23.0], [43.1, 54.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 126, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000549.mp4", "duration": 71.667, "tgt": [[11.4, 13.2], [17.7, 21.0], [22.8, 24.2], [31.7, 33.3], [54.8, 57.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 127, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000569.mp4", "duration": 88.167, "tgt": [[2.2, 3.3], [13.5, 14.4], [48.4, 49.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 128, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000569.mp4", "duration": 88.167, "tgt": [[2.7, 3.8], [49.3, 50.3], [65.0, 66.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 129, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000786.mp4", "duration": 98.0, "tgt": [[3.8, 5.9], [15.9, 20.3], [31.9, 34.1], [52.6, 54.7], [79.9, 82.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 130, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000786.mp4", "duration": 98.0, "tgt": [[33.3, 34.8], [53.5, 54.9], [81.8, 83.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 131, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000796.mp4", "duration": 127.333, "tgt": [[23.3, 24.4], [31.0, 31.9], [37.3, 38.2], [47.3, 48.3], [54.5, 55.6], [61.1, 62.1], [73.0, 74.3], [77.9, 78.8], [89.4, 90.7], [105.1, 106.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 132, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000798.mp4", "duration": 108.667, "tgt": [[87.4, 88.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 133, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000798.mp4", "duration": 108.667, "tgt": [[16.2, 19.8], [35.1, 38.9], [55.8, 59.3], [88.2, 89.6], [92.5, 100.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 134, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000897.mp4", "duration": 120.167, "tgt": [[10.4, 13.4], [17.6, 21.2], [33.3, 37.0], [52.8, 56.5], [65.0, 67.6], [88.6, 91.0], [97.6, 103.3], [103.4, 105.1], [105.2, 107.0]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 135, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000981.mp4", "duration": 28.0, "tgt": [[4.0, 5.0], [22.4, 25.6]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 136, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001313.mp4", "duration": 174.0, "tgt": [[21.1, 22.6], [36.1, 37.8], [162.2, 164.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 137, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001358.mp4", "duration": 219.5, "tgt": [[193.1, 194.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 138, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001358.mp4", "duration": 219.5, "tgt": [[0.1, 0.9], [1.2, 1.8], [2.1, 2.8], [129.5, 132.7], [160.8, 164.0], [166.3, 168.3], [180.8, 182.2], [186.0, 191.8], [194.0, 195.0], [199.6, 202.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 139, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001460.mp4", "duration": 29.0, "tgt": [[2.2, 3.2]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket bowling'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 140, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001460.mp4", "duration": 29.0, "tgt": [[2.9, 3.6], [14.2, 15.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 141, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000443.mp4", "duration": 291.333, "tgt": [[0.0, 7.1], [23.6, 33.4], [48.2, 55.8], [72.9, 79.8], [112.9, 119.8], [133.4, 139.8], [164.2, 171.9], [195.0, 199.9], [237.2, 245.4], [262.6, 270.3]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'hammer throw'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 142, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001072.mp4", "duration": 284.833, "tgt": [[6.8, 14.7], [38.2, 44.6], [62.4, 76.8], [93.3, 100.1], [116.9, 131.6], [154.3, 160.0], [181.4, 194.4], [247.3, 255.5]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'hammer throw'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 143, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0000448.mp4", "duration": 100.5, "tgt": [[16.0, 17.1], [59.0, 59.8], [62.9, 63.4], [66.1, 67.7]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 144, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001276.mp4", "duration": 62.333, "tgt": [[8.6, 10.4], [13.0, 15.3], [17.3, 19.3], [21.5, 23.4], [26.4, 27.8], [33.6, 35.5], [41.2, 43.0], [47.0, 48.4]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}, {"version": 1.0, "idx": 145, "task": "tal", "source": "thumos14", "video": "thumos14/video_test_0001325.mp4", "duration": 175.167, "tgt": [[36.0, 38.0], [47.8, 51.6], [92.4, 94.5], [150.4, 152.1]], "q": "You are given a video about human actions. Watch the video carefully and find all the visual events belonging to the action category: 'cricket shot'. The format of your response should be: 'The action happens in <start time> - <end time>, <start time> - <end time>, and <start time> - <end time>'."}]
annotations/vid/tal_thumos15.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/vid/tem_perception_test.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/vid/tem_qvhighlights.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/vid/tvg_charades_sta.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/vid/tvg_qvhighlights.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/vid/vhd_qvhighlights.json ADDED
The diff for this file is too large to render. See raw diff
 
annotations/vid/vhd_youtube_highlights.json ADDED
The diff for this file is too large to render. See raw diff
 
compress_videos.py ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import argparse
2
+ import multiprocessing as mp
3
+ import os
4
+ from functools import partial
5
+
6
+ import nncore
7
+
8
+
9
+ def _proc(blob, fps, size):
10
+ nncore.mkdir(nncore.dir_name(blob[1]))
11
+ cmd = f'ffmpeg -y -i {blob[0]} -filter:v "scale=\'if(gt(a,1),trunc(oh*a/2)*2,{size})\':\'if(gt(a,1),{size},trunc(ow*a/2)*2)\'" -map 0:v -r {fps} {blob[1]}'
12
+ nncore.exec(cmd)
13
+ assert nncore.is_file(blob[1]) and nncore.get_size(blob[1]) >= 1, blob
14
+
15
+
16
+ def parse_args():
17
+ parser = argparse.ArgumentParser()
18
+ parser.add_argument('--src_dir', default='videos')
19
+ parser.add_argument('--tgt_dir', default='videos_compressed')
20
+ parser.add_argument('--fps', type=int, default=3)
21
+ parser.add_argument('--size', type=int, default=224)
22
+ parser.add_argument('--workers', type=int)
23
+ args = parser.parse_args()
24
+ return args
25
+
26
+
27
+ if __name__ == '__main__':
28
+ args = parse_args()
29
+
30
+ assert args.src_dir != args.tgt_dir
31
+
32
+ blobs = []
33
+ for pattern in ('*.mp4', '*.mkv', '*.webp'):
34
+ for path in nncore.find(args.src_dir, pattern):
35
+ tgt_path = nncore.join(args.tgt_dir, f'{nncore.split_ext(path[len(args.src_dir) + 1:])[0]}.mp4')
36
+ blobs.append((path, tgt_path))
37
+
38
+ if len(blobs) == 0:
39
+ raise FileNotFoundError(f"No videos found in '{args.src_dir}'.")
40
+
41
+ print(f'Total number of videos: {len(blobs)}')
42
+ for b in blobs[:3]:
43
+ print(b)
44
+
45
+ print(f'Starting {os.cpu_count() if args.workers is None else args.workers} workers...')
46
+ proc = partial(_proc, fps=args.fps, size=args.size)
47
+
48
+ prog_bar = nncore.ProgressBar(num_tasks=len(blobs))
49
+ with mp.Pool(args.workers) as pool:
50
+ for _ in pool.imap_unordered(proc, blobs):
51
+ prog_bar.update()
evaluation/compute_metrics.py ADDED
@@ -0,0 +1,971 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # ---------------------------------------------
2
+ # Evaluation code for E.T. Bench
3
+ # Copyright (c) 2024 Ye Liu
4
+ # Licensed under CC BY-NC-SA 4.0 license
5
+ # ---------------------------------------------
6
+
7
+ import argparse
8
+ import copy
9
+ import random
10
+ import re
11
+ import string
12
+
13
+ import nncore
14
+ import numpy as np
15
+ import torch
16
+ from nncore.ops import temporal_iou
17
+ from tabulate import tabulate
18
+
19
+ import sentence_transformers
20
+ from pycocoevalcap.bleu.bleu import Bleu
21
+ from pycocoevalcap.cider.cider import Cider
22
+ from pycocoevalcap.meteor.meteor import Meteor
23
+ from pycocoevalcap.rouge.rouge import Rouge
24
+ from pycocoevalcap.tokenizer.ptbtokenizer import PTBTokenizer
25
+ from sentence_transformers.util import dot_score
26
+
27
+
28
+ class SentenceTransformerSimilarity(object):
29
+
30
+ def __init__(self):
31
+ self.model = sentence_transformers.SentenceTransformer('sentence-transformers/all-MiniLM-L6-v2')
32
+
33
+ def compute_sim(self, a, b):
34
+ a_emb = self.model.encode([a])
35
+ b_emb = self.model.encode([b])
36
+ score = dot_score(a_emb, b_emb)[0, 0].cpu()
37
+ return float(score)
38
+
39
+ def compute_score(self, a, b):
40
+ assert len(a) == len(b)
41
+ keys = list(a.keys())
42
+ aa, bb = [], []
43
+ for key in keys:
44
+ assert len(a[key]) == len(b[key]) == 1
45
+ aa.append(a[key][0])
46
+ bb.append(b[key][0])
47
+ a_emb = self.model.encode(aa)
48
+ b_emb = self.model.encode(bb)
49
+ score = dot_score(a_emb, b_emb).cpu()
50
+ assert score.shape[0] == score.shape[1]
51
+ score = [score[i, i].item() for i in range(score.shape[0])]
52
+ score = sum(score) / len(score)
53
+ return float(score), None
54
+
55
+
56
+ def random_string(string_length):
57
+ letters = string.ascii_lowercase
58
+ return ''.join(random.choice(letters) for i in range(string_length))
59
+
60
+
61
+ def remove_nonascii(text):
62
+ return ''.join([i if ord(i) < 128 else ' ' for i in text])
63
+
64
+
65
+ class DVCEval(object):
66
+
67
+ def __init__(self, ground_truth, prediction, tious=None, max_proposals=1000, sentsim=None):
68
+ self.tious = tious
69
+ self.max_proposals = max_proposals
70
+ self.ground_truths = [ground_truth]
71
+ self.prediction = self.import_prediction(prediction)
72
+ self.ground_truths_keys = list(ground_truth.keys())
73
+
74
+ self.tokenizer = PTBTokenizer(verbose=False)
75
+ self.scorers = [(Bleu(4), ['Bleu_1', 'Bleu_2', 'Bleu_3', 'Bleu_4']), (Meteor(), 'METEOR'), (Rouge(), 'ROUGE_L'),
76
+ (Cider(), 'CIDEr'), (sentsim, 'SentSim')]
77
+
78
+ def import_prediction(self, prediction):
79
+ results = dict()
80
+ for vid_id in prediction['results']:
81
+ results[vid_id] = prediction['results'][vid_id][:self.max_proposals]
82
+ return results
83
+
84
+ def iou(self, interval_1, interval_2):
85
+ start_i, end_i = interval_1[0], interval_1[1]
86
+ start, end = interval_2[0], interval_2[1]
87
+ intersection = max(0, min(end, end_i) - max(start, start_i))
88
+ union = min(max(end, end_i) - min(start, start_i), end - start + end_i - start_i)
89
+ iou = float(intersection) / (union + 1e-8)
90
+ return iou
91
+
92
+ def get_gt_vid_ids(self):
93
+ vid_ids = set([])
94
+ for gt in self.ground_truths:
95
+ vid_ids |= set(gt.keys())
96
+ return list(vid_ids)
97
+
98
+ def evaluate(self):
99
+ self.scores = dict()
100
+ for tiou in self.tious:
101
+ scores = self.evaluate_tiou(tiou)
102
+ for metric, score in scores.items():
103
+ if metric not in self.scores:
104
+ self.scores[metric] = []
105
+ self.scores[metric].append(score)
106
+ self.scores['Recall'] = []
107
+ self.scores['Precision'] = []
108
+ for tiou in self.tious:
109
+ precision, recall = self.evaluate_detection(tiou)
110
+ self.scores['Recall'].append(recall)
111
+ self.scores['Precision'].append(precision)
112
+
113
+ def evaluate_detection(self, tiou):
114
+ gt_vid_ids = self.get_gt_vid_ids()
115
+ recall = [0] * len(gt_vid_ids)
116
+ precision = [0] * len(gt_vid_ids)
117
+ for vid_i, vid_id in enumerate(gt_vid_ids):
118
+ best_recall = 0
119
+ best_precision = 0
120
+ for gt in self.ground_truths:
121
+ if vid_id not in gt:
122
+ continue
123
+ refs = gt[vid_id]
124
+ ref_set_covered = set([])
125
+ pred_set_covered = set([])
126
+ if vid_id in self.prediction:
127
+ for pred_i, pred in enumerate(self.prediction[vid_id]):
128
+ pred_timestamp = pred['timestamp']
129
+ for ref_i, ref_timestamp in enumerate(refs['timestamps']):
130
+ if self.iou(pred_timestamp, ref_timestamp) > tiou:
131
+ ref_set_covered.add(ref_i)
132
+ pred_set_covered.add(pred_i)
133
+
134
+ new_precision = float(len(pred_set_covered)) / (pred_i + 1)
135
+ best_precision = max(best_precision, new_precision)
136
+ new_recall = float(len(ref_set_covered)) / len(refs['timestamps'])
137
+ best_recall = max(best_recall, new_recall)
138
+ recall[vid_i] = best_recall
139
+ precision[vid_i] = best_precision
140
+ return sum(precision) / len(precision), sum(recall) / len(recall)
141
+
142
+ def evaluate_tiou(self, tiou):
143
+ vid2capid, res, gts, cur_res, cur_gts = dict(), dict(), dict(), dict(), dict()
144
+ unique_index = 0
145
+ gt_vid_ids = self.get_gt_vid_ids()
146
+
147
+ for vid_id in gt_vid_ids:
148
+ vid2capid[vid_id] = []
149
+
150
+ if vid_id not in self.prediction:
151
+ pass
152
+
153
+ else:
154
+ for pred in self.prediction[vid_id]:
155
+ has_added = False
156
+ for gt in self.ground_truths:
157
+ if vid_id not in gt:
158
+ continue
159
+ gt_captions = gt[vid_id]
160
+ for caption_idx, caption_timestamp in enumerate(gt_captions['timestamps']):
161
+ if self.iou(pred['timestamp'], caption_timestamp) >= tiou:
162
+ cur_res[unique_index] = [{'caption': remove_nonascii(pred['sentence'])}]
163
+ cur_gts[unique_index] = [{
164
+ 'caption':
165
+ remove_nonascii(gt_captions['sentences'][caption_idx])
166
+ }]
167
+ vid2capid[vid_id].append(unique_index)
168
+ unique_index += 1
169
+ has_added = True
170
+
171
+ if not has_added:
172
+ cur_res[unique_index] = [{'caption': remove_nonascii(pred['sentence'])}]
173
+ cur_gts[unique_index] = [{'caption': random_string(random.randint(10, 20))}]
174
+ vid2capid[vid_id].append(unique_index)
175
+ unique_index += 1
176
+
177
+ output = dict()
178
+ for scorer, method in self.scorers:
179
+ all_scores = dict()
180
+
181
+ tokenize_res = self.tokenizer.tokenize(cur_res)
182
+ tokenize_gts = self.tokenizer.tokenize(cur_gts)
183
+
184
+ for vid in vid2capid.keys():
185
+ res[vid] = {index: tokenize_res[index] for index in vid2capid[vid]}
186
+ gts[vid] = {index: tokenize_gts[index] for index in vid2capid[vid]}
187
+
188
+ for vid_id in gt_vid_ids:
189
+ if len(res[vid_id]) == 0 or len(gts[vid_id]) == 0:
190
+ if isinstance(method, list):
191
+ score = [0] * len(method)
192
+ else:
193
+ score = 0
194
+ else:
195
+ if isinstance(method, list):
196
+ score, scores = scorer.compute_score(gts[vid_id], res[vid_id], verbose=0)
197
+ else:
198
+ score, scores = scorer.compute_score(gts[vid_id], res[vid_id])
199
+
200
+ all_scores[vid_id] = score
201
+
202
+ if isinstance(method, list):
203
+ scores = np.mean(list(all_scores.values()), axis=0)
204
+ for m in range(len(method)):
205
+ output[method[m]] = scores[m]
206
+ else:
207
+ output[method] = np.mean(list(all_scores.values()))
208
+
209
+ return output
210
+
211
+
212
+ def extract_time_part(time_part):
213
+ radius = 20
214
+ extracted_time_part = re.compile(r"\d+\.*\d*\s*-\s*\d+\.*\d*").findall(time_part)
215
+
216
+ if len(extracted_time_part) == 0:
217
+ if time_part.count(':') == 1:
218
+ extracted_time = re.compile(r"\d+\.*\d*:\d+\.*\d*").findall(time_part)[0]
219
+ extracted_time = int(extracted_time.split(':')[0]) * 60 + int(extracted_time.split(':')[1])
220
+ if extracted_time > radius:
221
+ extracted_time_part = [f'{extracted_time - radius} - {extracted_time + radius}']
222
+ else:
223
+ extracted_time_part = [f'{extracted_time} - {extracted_time + 2*radius}']
224
+ elif time_part.count(':') == 2:
225
+ start, end = re.compile(r"\d+\.*\d*:\d+\.*\d*").findall(time_part)
226
+ start_seconds = int(start.split(':')[0]) * 60 + int(start.split(':')[1])
227
+ end_seconds = int(end.split(':')[0]) * 60 + int(end.split(':')[1])
228
+ extracted_time_part = [f'{start_seconds} - {end_seconds}']
229
+
230
+ if len(extracted_time_part) == 0:
231
+ extracted_time_part = re.compile(r"\d+\.*\d*(?!\.)").findall(time_part)
232
+ if len(extracted_time_part) == 1:
233
+ extracted_time = float(extracted_time_part[0])
234
+ if extracted_time > radius:
235
+ extracted_time_part = [f'{extracted_time - radius} - {extracted_time + radius}']
236
+ else:
237
+ extracted_time_part = [f'{extracted_time} - {extracted_time + 2 * radius}']
238
+ elif len(extracted_time_part) == 2:
239
+ extracted_time_part = [f'{extracted_time_part[0]} - {extracted_time_part[1]}']
240
+
241
+ return extracted_time_part
242
+
243
+
244
+ def extract_time_from_paragraph(paragraph):
245
+ paragraph = paragraph.lower()
246
+ patterns = [(r"(\d+\.*\d*)\s*-\s*(\d+\.*\d*)", r"(\d+\.*\d*\s*-\s*\d+\.*\d*)")]
247
+ timestamps, captions = [], []
248
+
249
+ for time_pattern, string_pattern in patterns:
250
+ time_matches = re.findall(time_pattern, paragraph)
251
+ string_matches = re.findall(string_pattern, paragraph)
252
+
253
+ if time_matches:
254
+ timestamps = [[float(start), float(end)] for start, end in time_matches]
255
+ rest_para = paragraph
256
+ for time_string in string_matches:
257
+ rest_para = rest_para.replace(time_string, '\n')
258
+ captions = rest_para.replace('seconds', '').split('\n')
259
+ if len(timestamps) > 0:
260
+ break
261
+
262
+ if len(timestamps) == 0:
263
+ start_time_pattern = r"(?:start(?:ing)? time: (\d+\.*\d*)(?:s| seconds)?)"
264
+ end_time_pattern = r"(?:end(?:ing)? time: (\d+\.*\d*)(?:s| seconds)?)"
265
+ end_matches = re.findall(end_time_pattern, paragraph, re.DOTALL | re.IGNORECASE)
266
+ start_matches = re.findall(start_time_pattern, paragraph, re.DOTALL | re.IGNORECASE)
267
+
268
+ if start_matches and end_matches:
269
+ timestamps = [[float(start), float(end)] for start, end in zip(start_matches, end_matches)]
270
+ captions = re.findall(r"description: (.*)", paragraph)
271
+ if len(captions) == 0:
272
+ captions = re.findall(r"\*\s*(.*)", paragraph)
273
+
274
+ if len(timestamps) == 0:
275
+ start_end_matches = re.findall(r"start time (\d+\.*\d*), end time (\d+\.*\d*)", paragraph)
276
+ if start_end_matches:
277
+ timestamps = list(start_end_matches)
278
+ for (start, end) in start_end_matches:
279
+ paragraph = paragraph.replace(f'start time {start}, end time {end}', '\n')
280
+ captions = paragraph.split('\n')
281
+ assert len(timestamps) <= 0
282
+
283
+ captions = [c.strip().strip(', ').rstrip() for c in captions if len(c) > 5]
284
+ min_len = min(len(timestamps), len(captions))
285
+ timestamps = timestamps[:min_len]
286
+ captions = captions[:min_len]
287
+
288
+ assert len(timestamps) == len(captions)
289
+ return timestamps, captions
290
+
291
+
292
+ def dvc_format(caption):
293
+ timestamps = []
294
+ sents = []
295
+ paras = caption
296
+
297
+ try:
298
+ timestamps, sents = extract_time_from_paragraph(paras)
299
+ except Exception:
300
+ return None, None
301
+
302
+ if len(timestamps) == 0:
303
+ if '\n' in caption:
304
+ caps = caption.split('\n')
305
+ caps = [c for c in caps if len(c) > 7]
306
+ else:
307
+ raw_caps = caption.split('.')
308
+ caps = [c for c in raw_caps if len(c) > 7]
309
+ caps = [c + '.' for c in caps]
310
+ for cap in caps:
311
+ try:
312
+ if len(timestamps) == 0:
313
+ parts = cap.split('seconds')
314
+ parts = [p.strip(',') for p in parts]
315
+ time_part = parts[0]
316
+ extracted_time_part = extract_time_part(time_part)
317
+ if len(extracted_time_part) == 0:
318
+ continue
319
+ else:
320
+ time_part = extracted_time_part[0]
321
+ sent_part = parts[-1]
322
+ stime = round(float(time_part.split('-')[0].strip()), 2)
323
+ etime = round(float(time_part.split('-')[1].strip()), 2)
324
+ timestamps.append([stime, etime])
325
+ sents.append(sent_part.strip())
326
+ except Exception:
327
+ continue
328
+
329
+ assert len(timestamps) == len(sents)
330
+
331
+ if len(timestamps) == 0:
332
+ return None, None
333
+
334
+ for i in range(len(timestamps)):
335
+ assert isinstance(timestamps[i], list) and len(timestamps[i]) == 2 and isinstance(
336
+ timestamps[i][0], (int, float)) and isinstance(timestamps[i][1], (int, float))
337
+ timestamps[i] = [min(timestamps[i]), max(timestamps[i])]
338
+
339
+ return timestamps, sents
340
+
341
+
342
+ def tvg_format(ans):
343
+ ans = ans.lower()
344
+ sentences = re.split(r'[!?\n]', ans)
345
+
346
+ keywords = ['starts', 'ends', 'happens in', 'start time', 'end time', 'start', 'end', 'happen']
347
+ candidates = []
348
+ for sentence in sentences:
349
+ if any(keyword in sentence for keyword in keywords):
350
+ candidates.append(sentence)
351
+
352
+ timestamps = []
353
+ patterns = [r"(\d+\.*\d*)\s*-\s*(\d+\.*\d*)"]
354
+
355
+ for time_pattern in patterns:
356
+ time_matches = re.findall(time_pattern, ans)
357
+ if time_matches:
358
+ timestamps = [[float(start), float(end)] for start, end in time_matches]
359
+
360
+ if len(timestamps) == 0:
361
+ pattern = r"(\d+\.*\d*)\s* to \s*(\d+\.*\d*)"
362
+ time_matches = re.findall(pattern, ans)
363
+ if time_matches:
364
+ timestamps = [[float(start), float(end)] for start, end in time_matches]
365
+
366
+ if len(sentences) == 0:
367
+ return None
368
+
369
+ if len(timestamps) == 0:
370
+ times = []
371
+ time_regex = re.compile(r'\b(\d+\.\d+\b|\b\d+)\b')
372
+ for sentence in candidates:
373
+ time = re.findall(time_regex, sentence)
374
+ if time:
375
+ time_in_sec = float(time[0])
376
+ times.append(time_in_sec)
377
+ times = times[:len(times) // 2 * 2]
378
+ timestamps = [(times[i], times[i + 1]) for i in range(0, len(times), 2)]
379
+ if len(timestamps) == 0:
380
+ times = []
381
+ time_regex = re.compile(r'\b((\d{1,2}:\d{2}:\d{2}))\b')
382
+ for sentence in candidates:
383
+ time = re.findall(time_regex, sentence)
384
+ if time:
385
+ t = time[0]
386
+ else:
387
+ continue
388
+ if t.count(':') == 2:
389
+ h, m, s = map(int, t.split(':'))
390
+ time_in_sec = h * 3600 + m * 60 + s
391
+ elif t.count(':') == 1:
392
+ m, s = map(int, t.split(':'))
393
+ time_in_sec = m * 60 + s
394
+ times.append(time_in_sec)
395
+ times = times[:len(times) // 2 * 2]
396
+ timestamps = [(times[i], times[i + 1]) for i in range(0, len(times), 2)]
397
+ results = []
398
+ for (start, end) in timestamps:
399
+ if end > start:
400
+ results.append([start, end])
401
+ else:
402
+ results.append([end, start])
403
+
404
+ if len(results) == 0:
405
+ results = None
406
+
407
+ if results is not None:
408
+ assert isinstance(results, list)
409
+ for item in results:
410
+ assert isinstance(item, list)
411
+ assert len(item) == 2
412
+ assert isinstance(item[0], (int, float))
413
+ assert isinstance(item[1], (int, float))
414
+
415
+ return results
416
+
417
+
418
+ def tvg_eval(samples):
419
+ iou_thr = [0.1, 0.3, 0.5, 0.7]
420
+ hit = [0 for _ in iou_thr]
421
+ cnt, sum_iou = 0, 0
422
+ for sample in samples:
423
+ gt = sample['tgt']
424
+
425
+ pred = tvg_format(sample['a'])
426
+ if pred is None:
427
+ cnt += 1
428
+ continue
429
+
430
+ pred = pred[0]
431
+ gt = torch.Tensor([gt])
432
+ pred = torch.Tensor([pred])
433
+ iou = temporal_iou(gt, pred).item()
434
+ sum_iou += iou
435
+
436
+ for i, thr in enumerate(iou_thr):
437
+ if iou >= thr:
438
+ hit[i] += 1
439
+
440
+ recall = [h / len(samples) for h in hit]
441
+ miou = sum_iou / len(samples)
442
+
443
+ out = dict(Total=len(samples), Failed=cnt, mIoU=round(miou, 5))
444
+ for rec, thr in zip(recall, iou_thr):
445
+ out[f'F1@{thr}'] = round(rec, 5)
446
+ out['F1'] = round(sum(recall) / len(recall), 5)
447
+ return out
448
+
449
+
450
+ def vhd_eval(samples):
451
+ hit, cnt = 0, 0
452
+ for sample in samples:
453
+ gt = sample['tgt']
454
+ if not isinstance(gt[0][0], (list, tuple)):
455
+ gt = [gt]
456
+
457
+ match = re.search(r"[-+]?\d*\.\d+|[-+]?\d+", sample['a'])
458
+ if not match:
459
+ cnt += 1
460
+ continue
461
+
462
+ pred = float(match.group(0))
463
+ matched = False
464
+ for annotator in gt:
465
+ for g in annotator:
466
+ if pred >= g[0] and pred <= g[1]:
467
+ matched = True
468
+ break
469
+ if matched:
470
+ hit += 1
471
+
472
+ out = dict(Total=len(samples), Failed=cnt)
473
+ out['F1'] = round(hit / len(samples), 5)
474
+ return out
475
+
476
+
477
+ def tem_eval(samples):
478
+ iou_thr = [0.1, 0.3, 0.5, 0.7]
479
+ hit = [0 for _ in iou_thr]
480
+ cnt, sum_iou = 0, 0
481
+ for sample in samples:
482
+ gt = sample['tgt']
483
+
484
+ pred = tvg_format(sample['a'])
485
+ if pred is None:
486
+ cnt += 1
487
+ continue
488
+
489
+ pred = pred[0]
490
+ gt = torch.Tensor(gt)
491
+ pred = torch.Tensor([pred])
492
+ iou = temporal_iou(gt, pred).max().item()
493
+ sum_iou += iou
494
+
495
+ for i, thr in enumerate(iou_thr):
496
+ if iou >= thr:
497
+ hit[i] += 1
498
+
499
+ recall = [h / len(samples) for h in hit]
500
+ miou = sum_iou / len(samples)
501
+
502
+ out = dict(Total=len(samples), Failed=cnt, mIoU=round(miou, 5))
503
+ for rec, thr in zip(recall, iou_thr):
504
+ out[f'R@{thr}'] = round(rec, 5)
505
+ out['mRec'] = round(sum(recall) / len(recall), 5)
506
+ return out
507
+
508
+
509
+ def tal_eval(samples):
510
+ iou_thr = [0.1, 0.3, 0.5, 0.7]
511
+ f1_score = [0 for _ in iou_thr]
512
+ cnt = 0
513
+ for sample in samples:
514
+ gt = sample['tgt']
515
+
516
+ pred = tvg_format(sample['a'])
517
+ if pred is None:
518
+ cnt += 1
519
+ continue
520
+
521
+ gt = torch.Tensor(gt)
522
+ pred = torch.Tensor(pred)
523
+ iou = temporal_iou(gt, pred)
524
+
525
+ for i, thr in enumerate(iou_thr):
526
+ if iou.max() < thr:
527
+ continue
528
+ else:
529
+ rec = (iou.amax(dim=1) >= thr).float().mean().item()
530
+ prc = (iou.amax(dim=0) >= thr).float().mean().item()
531
+ f1_score[i] += 2 * prc * rec / (prc + rec)
532
+
533
+ f1_score = [f / len(samples) for f in f1_score]
534
+
535
+ out = dict(Total=len(samples), Failed=cnt)
536
+ for f1, thr in zip(f1_score, iou_thr):
537
+ out[f'F1@{thr}'] = round(f1, 5)
538
+ out['F1'] = round(sum(f1_score) / len(f1_score), 5)
539
+ return out
540
+
541
+
542
+ def evs_eval(samples):
543
+ f1_score = []
544
+ cnt = 0
545
+ for sample in samples:
546
+ gt = sample['tgt']
547
+
548
+ pred = tvg_format(sample['a'])
549
+ if pred is None:
550
+ cnt += 1
551
+ continue
552
+
553
+ gt_map = torch.zeros(1000)
554
+ gt_len = 0
555
+ for g in gt:
556
+ s = max(0, round(g[0]))
557
+ e = round(g[1])
558
+ gt_map[s:e] = 1
559
+ gt_len += e - s
560
+
561
+ pred_map = torch.zeros(1000)
562
+ pred_len = 0
563
+ for p in pred:
564
+ s = max(0, round(p[0]))
565
+ e = round(p[1])
566
+ pred_map[s:e] = 2
567
+ pred_len += e - s
568
+
569
+ com_map = gt_map + pred_map
570
+
571
+ tp = (com_map == 3).sum().item()
572
+ fp = (com_map == 2).sum().item()
573
+ fn = (com_map == 1).sum().item()
574
+
575
+ if tp == 0:
576
+ f1 = 0
577
+ else:
578
+ rec = tp / (tp + fn)
579
+ prc = tp / (tp + fp)
580
+ f1 = 2 * prc * rec / (prc + rec)
581
+
582
+ f1_score.append(f1)
583
+
584
+ f1_score = round(sum(f1_score) / len(f1_score), 5) if len(f1_score) > 0 else 0
585
+
586
+ out = dict(Total=len(samples), Failed=cnt)
587
+ out['F1'] = f1_score
588
+ return out
589
+
590
+
591
+ def rvq_eval(samples, st):
592
+ if len(samples[0]['o']) == 4:
593
+ match_map = dict(a=0, b=1, c=2, d=3)
594
+ elif len(samples[0]['o']) == 5:
595
+ match_map = dict(a=0, b=1, c=2, d=3, e=4)
596
+ else:
597
+ raise NotImplementedError
598
+
599
+ hit, cnt = 0, 0
600
+ for sample in samples:
601
+ gt = sample['p']
602
+ pred = sample['a']
603
+
604
+ ever_matched = False
605
+ match = re.search(r'\(([A-Za-z])\)', pred)
606
+ if match:
607
+ ever_matched = True
608
+ choice = match.group(1).lower()
609
+ if choice in match_map and gt == match_map[choice]:
610
+ hit += 1
611
+ continue
612
+
613
+ pred = pred.lower()
614
+ if pred.startswith('best option:'):
615
+ pred = pred[12:]
616
+
617
+ pred = pred.lstrip().lstrip('(').lstrip()
618
+ if len(pred) == 0:
619
+ cnt += 1
620
+ continue
621
+
622
+ if len(pred) == 1 or pred[1] in ('.', ',', ' ', ')'):
623
+ ever_matched = True
624
+ if pred[0] in match_map and gt == match_map[pred[0]]:
625
+ hit += 1
626
+ continue
627
+
628
+ hit_idx, max_score = 0, float('-inf')
629
+ _map = ['A', 'B', 'C', 'D', 'E']
630
+ for idx, option in enumerate(sample['o']):
631
+ if isinstance(option, (list, tuple)):
632
+ opt = f'{option[0]} - {option[1]}'
633
+ else:
634
+ opt = option
635
+ opt = f'({_map[idx]}) {opt}'
636
+ sim = st.compute_sim(pred, opt)
637
+ if sim > max_score:
638
+ hit_idx = idx
639
+ max_score = sim
640
+
641
+ assert max_score != float('-inf')
642
+ if not ever_matched:
643
+ cnt += 1
644
+
645
+ if gt == hit_idx:
646
+ hit += 1
647
+
648
+ acc = hit / len(samples)
649
+ out = dict(Total=len(samples), Failed=cnt, Acc=round(acc, 5))
650
+ return out
651
+
652
+
653
+ def gvq_eval(samples, st):
654
+ acc_hit_idx, acc_cnt = [], 0
655
+ _samples = copy.deepcopy(samples)
656
+ for sample_idx, sample in enumerate(_samples):
657
+ gt = sample['p']
658
+ pred = sample['a']
659
+
660
+ if pred.lower().startswith('best option:'):
661
+ pred = pred[12:]
662
+
663
+ pred = pred.lstrip().lstrip('(').lstrip()
664
+ if len(pred) == 0:
665
+ acc_cnt += 1
666
+ continue
667
+
668
+ if len(sample['o']) == 4:
669
+ match_map = dict(a=0, b=1, c=2, d=4)
670
+ elif len(sample['o']) == 5:
671
+ match_map = dict(a=0, b=1, c=2, d=4, e=5)
672
+ else:
673
+ raise NotImplementedError
674
+
675
+ if len(pred) == 1 or pred[1] in ('.', ',', ' ', ')'):
676
+ if pred[0].lower() in match_map:
677
+ if gt == match_map[pred[0].lower()]:
678
+ acc_hit_idx.append(sample_idx)
679
+ continue
680
+
681
+ hit_idx, max_score = 0, float('-inf')
682
+ _map = ['A', 'B', 'C', 'D', 'E']
683
+ for idx, option in enumerate(sample['o']):
684
+ if isinstance(option, (list, tuple)):
685
+ opt = f'{option[0]} - {option[1]}'
686
+ else:
687
+ opt = option
688
+ opt = f'({_map[idx]}) {opt}'
689
+ sim = st.compute_sim(pred, opt)
690
+ if sim > max_score:
691
+ hit_idx = idx
692
+ max_score = sim
693
+ if max_score == float('-inf'):
694
+ acc_cnt += 1
695
+ continue
696
+
697
+ if gt == hit_idx:
698
+ acc_hit_idx.append(sample_idx)
699
+
700
+ acc_hit_idx = set(acc_hit_idx)
701
+ iou_thr = [0.1, 0.3, 0.5, 0.7]
702
+ hit = [0 for _ in iou_thr]
703
+ rec_cnt, sum_iou = 0, 0
704
+ for sample_idx, sample in enumerate(samples):
705
+ if sample_idx not in acc_hit_idx:
706
+ continue
707
+
708
+ gt = sample['tgt']
709
+
710
+ pred = tvg_format(sample['a'])
711
+ if pred is None:
712
+ rec_cnt += 1
713
+ continue
714
+
715
+ pred = pred[0]
716
+ gt = torch.Tensor([gt])
717
+ pred = torch.Tensor([pred])
718
+ iou = temporal_iou(gt, pred).item()
719
+ sum_iou += iou
720
+
721
+ for i, thr in enumerate(iou_thr):
722
+ if iou >= thr:
723
+ hit[i] += 1
724
+
725
+ recall = [h / len(samples) for h in hit]
726
+ miou = sum_iou / len(samples)
727
+
728
+ out = dict(Total=len(samples), Failed=rec_cnt + acc_cnt, mIoU=round(miou, 5))
729
+ for rec, thr in zip(recall, iou_thr):
730
+ out[f'R@{thr}'] = round(rec, 5)
731
+ out['mRec'] = round(sum(recall) / len(recall), 5)
732
+ out['Acc'] = round(len(acc_hit_idx) / len(samples), 5)
733
+ return out
734
+
735
+
736
+ def dvc_eval(samples, st):
737
+ iou_thr = [0.1, 0.3, 0.5, 0.7]
738
+ gt_dict, pred = dict(), dict(results=dict())
739
+ cnt = 0
740
+ for sample in samples:
741
+ gt = sample['tgt']
742
+ gt_cap = sample['g']
743
+
744
+ time, cap = dvc_format(sample['a'])
745
+ if time is None or cap is None:
746
+ cnt += 1
747
+ continue
748
+
749
+ gt_dict[sample['video']] = dict(timestamps=gt, sentences=gt_cap)
750
+ pred['results'][sample['video']] = [dict(sentence=c, timestamp=t) for t, c in zip(time, cap)]
751
+
752
+ scale = len(pred['results']) / len(samples)
753
+
754
+ if gt_dict:
755
+ evaluator = DVCEval(ground_truth=gt_dict, prediction=pred, tious=iou_thr, sentsim=st)
756
+ evaluator.evaluate()
757
+ scores = evaluator.scores
758
+ else:
759
+ scores = dict()
760
+ for key in ('Recall', 'Precision', 'Bleu_1', 'Bleu_2', 'Bleu_3', 'Bleu_4', 'METEOR', 'ROUGE_L', 'CIDEr',
761
+ 'SentSim'):
762
+ scores[key] = [0] * len(iou_thr)
763
+
764
+ out = dict(Total=len(samples), Failed=cnt)
765
+ f1_score = []
766
+ for rec, prc, thr in zip(scores['Recall'], scores['Precision'], iou_thr):
767
+ rec = rec * scale
768
+ prc = prc * scale
769
+ f1 = 0 if prc + rec == 0 else 2 * prc * rec / (prc + rec)
770
+ out[f'F1@{thr}'] = round(f1, 5)
771
+ f1_score.append(f1)
772
+ out['F1'] = round(sum(f1_score) / len(f1_score), 5)
773
+ for key in ('Bleu_1', 'Bleu_2', 'Bleu_3', 'Bleu_4', 'METEOR', 'ROUGE_L', 'CIDEr', 'SentSim'):
774
+ out[key] = round(sum(scores[key]) / len(scores[key]), 5)
775
+ return out
776
+
777
+
778
+ def parse_args():
779
+ parser = argparse.ArgumentParser()
780
+ parser.add_argument('pred_path')
781
+ parser.add_argument('--subset', action='store_true')
782
+ args = parser.parse_args()
783
+ return args
784
+
785
+
786
+ if __name__ == '__main__':
787
+ args = parse_args()
788
+
789
+ if args.pred_path.endswith('.json') or args.pred_path.endswith('.jsonl'):
790
+ pred_paths = [args.pred_path]
791
+ dir_name = nncore.dir_name(args.pred_path)
792
+ else:
793
+ pred_paths = nncore.ls(args.pred_path, ext=['json', 'jsonl'], join_path=True)
794
+ pred_paths = [path for path in pred_paths if path != 'metrics.json']
795
+ dir_name = args.pred_path
796
+
797
+ log_file = nncore.join(dir_name, 'metrics.log')
798
+ nncore.set_default_logger('etbench', fmt=None, log_file=log_file)
799
+
800
+ nncore.log(f'Total number of files: {len(pred_paths)}')
801
+
802
+ all_samples = []
803
+ for path in pred_paths:
804
+ nncore.log(f'Loading {path}...')
805
+ all_samples += nncore.load(path)
806
+
807
+ nncore.log(f'Total number of samples: {len(all_samples)}')
808
+
809
+ if args.subset:
810
+ subset = nncore.load(nncore.same_dir(__file__, 'subset.json'))
811
+
812
+ pred = dict()
813
+ for sample in all_samples:
814
+ task, source, idx = sample['task'], sample['source'], sample['idx']
815
+
816
+ if task not in pred:
817
+ pred[task] = dict()
818
+
819
+ if source not in pred[task]:
820
+ pred[task][source] = []
821
+
822
+ if not args.subset or (source in subset[task] and idx in subset[task][source]):
823
+ pred[task][source].append(sample)
824
+
825
+ if args.subset:
826
+ cnt = sum(len(d) for t in pred.values() for d in t.values())
827
+ nncore.log(f'Evaluating on a subset with {cnt} samples')
828
+
829
+ print('==========================================')
830
+ print('Start evaluation...')
831
+
832
+ st = SentenceTransformerSimilarity()
833
+
834
+ collected = dict()
835
+ for task in pred:
836
+ for source in pred[task]:
837
+ print(f'{task}_{source}: {len(pred[task][source])}')
838
+ if task in ('tvg', 'epm'):
839
+ out = tvg_eval(pred[task][source])
840
+ elif task in ('vhd', ):
841
+ out = vhd_eval(pred[task][source])
842
+ elif task in ('tem', ):
843
+ out = tem_eval(pred[task][source])
844
+ elif task in ('tal', ):
845
+ out = tal_eval(pred[task][source])
846
+ elif task in ('evs', ):
847
+ out = evs_eval(pred[task][source])
848
+ elif task in ('dvc', 'slc'):
849
+ out = dvc_eval(pred[task][source], st)
850
+ elif task in ('rar', 'rvq', 'eca'):
851
+ out = rvq_eval(pred[task][source], st)
852
+ elif task in ('gvq', ):
853
+ out = gvq_eval(pred[task][source], st)
854
+ else:
855
+ raise NotImplementedError
856
+
857
+ if task not in collected:
858
+ collected[task] = dict()
859
+
860
+ collected[task][source] = out
861
+
862
+ nncore.log('==========================================')
863
+
864
+ met = []
865
+
866
+ tasks = ['rar', 'eca', 'rvq']
867
+ if any(t in collected for t in tasks):
868
+ nncore.log('\nReferring\n')
869
+ out = [('Task', 'Source', 'Total', 'Failed', 'Acc')]
870
+ mean_acc = []
871
+ for task in tasks:
872
+ if task not in collected:
873
+ continue
874
+ task_acc = []
875
+ for source in collected[task]:
876
+ d = collected[task][source]
877
+ out.append((task, source, d['Total'], d['Failed'], d['Acc']))
878
+ mean_acc.append(d['Acc'])
879
+ task_acc.append(d['Acc'])
880
+ met.append(round(sum(task_acc) / len(task_acc), 5))
881
+ nncore.log(tabulate(out))
882
+ nncore.log(f'Mean Acc: {round(sum(mean_acc) / len(mean_acc), 5)}')
883
+
884
+ tasks = ['tvg', 'epm', 'tal', 'evs', 'vhd']
885
+ if any(t in collected for t in tasks):
886
+ nncore.log('\nGrounding\n')
887
+ out = [('Task', 'Source', 'Total', 'Failed', '[email protected]', '[email protected]', '[email protected]', '[email protected]', 'F1')]
888
+ mean_rec = []
889
+ for task in tasks:
890
+ if task not in collected:
891
+ continue
892
+ task_rec = []
893
+ for source in collected[task]:
894
+ d = collected[task][source]
895
+ o = [task, source, d['Total'], d['Failed']]
896
+ for thr in [0.1, 0.3, 0.5, 0.7]:
897
+ o.append(d.get(f'F1@{thr}', '-'))
898
+ o.append(d.get('F1', '-'))
899
+ out.append(tuple(o))
900
+ mean_rec.append(o[-1])
901
+ task_rec.append(o[-1])
902
+ met.append(round(sum(task_rec) / len(task_rec), 5))
903
+ nncore.log(tabulate(out))
904
+ nncore.log(f'Mean F1: {round(sum(mean_rec) / len(mean_rec), 5)}')
905
+
906
+ tasks = ['dvc', 'slc']
907
+ if any(t in collected for t in tasks):
908
+ nncore.log('\nCaptioning\n')
909
+ out = [('Task', 'Source', 'Total', 'Failed', '[email protected]', '[email protected]', '[email protected]', '[email protected]', 'F1', 'METEOR', 'ROUGE_L',
910
+ 'CIDEr', 'SentSim')]
911
+ mean_rec, mean_sim = [], []
912
+ for task in tasks:
913
+ if task not in collected:
914
+ continue
915
+ task_rec, task_sim = [], []
916
+ for source in collected[task]:
917
+ d = collected[task][source]
918
+ o = [task, source]
919
+ for key in out[0][2:]:
920
+ o.append(d[key])
921
+ out.append(tuple(o))
922
+ mean_rec.append(d['F1'])
923
+ task_rec.append(d['F1'])
924
+ mean_sim.append(d['SentSim'])
925
+ task_sim.append(d['SentSim'])
926
+ met.append(round(sum(task_rec) / len(task_rec), 5))
927
+ met.append(round(sum(task_sim) / len(task_sim), 5))
928
+ nncore.log(tabulate(out))
929
+ nncore.log(f'Mean F1: {round(sum(mean_rec) / len(mean_rec), 5)}')
930
+ nncore.log(f'Mean SentSim: {round(sum(mean_sim) / len(mean_sim), 5)}')
931
+
932
+ tasks = ['tem', 'gvq']
933
+ if any(t in collected for t in tasks):
934
+ nncore.log('\nComplex\n')
935
+ out = [('Task', 'Source', 'Total', 'Failed', '[email protected]', '[email protected]', '[email protected]', '[email protected]', 'mRec', 'Acc')]
936
+ for task in tasks:
937
+ if task not in collected:
938
+ continue
939
+ task_rec = []
940
+ for source in collected[task]:
941
+ d = collected[task][source]
942
+ o = [task, source]
943
+ for key in out[0][2:]:
944
+ o.append(d.get(key, '-'))
945
+ out.append(tuple(o))
946
+ task_rec.append(o[-2])
947
+ met.append(round(sum(task_rec) / len(task_rec), 5))
948
+ nncore.log(tabulate(out))
949
+ nncore.log(f'TEM Mean Rec: {met[-2]}')
950
+ nncore.log(f'GVQ Mean Rec: {met[-1]}')
951
+
952
+ nncore.log('\nOverall\n')
953
+ out = [('RAR (Acc)', 'EVC (Acc)', 'RVQ (Acc)', 'TVG (F1)', 'EPM (F1)', 'TAL (F1)', 'EVS (F1)', 'VHD (F1)',
954
+ 'DVC (F1)', 'DVC (Sim)', 'SLC (F1)', 'SLC (Sim)', 'TEM (Rec)', 'GVQ (Rec)')]
955
+ out.append(tuple([str(round(m * 100, 1)) for m in met]))
956
+ nncore.log(tabulate(out))
957
+
958
+ nncore.log('\nMerged\n')
959
+ out = [('Acc (ref)', 'F1 (gnd)', 'F1 (cap)', 'Sim (cap)', 'Rec (com)')]
960
+ out.append(
961
+ tuple([
962
+ str(round(sum(met[:3]) * 100 / 3, 1)),
963
+ str(round(sum(met[3:8]) * 100 / 5, 1)),
964
+ str(round((met[8] + met[10]) * 100 / 2, 1)),
965
+ str(round((met[9] + met[11]) * 100 / 2, 1)),
966
+ str(round(sum(met[12:]) * 100 / 2, 1))
967
+ ]))
968
+ nncore.log(tabulate(out))
969
+
970
+ path = nncore.join(dir_name, 'metrics.json')
971
+ nncore.dump(collected, path, indent=4)
evaluation/requirements.txt ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ git+https://github.com/salaniz/pycocoevalcap.git@a24f74c
2
+ nncore==0.4.5
3
+ numpy==1.26.4
4
+ sentence-transformers==3.0.1
5
+ torch==2.4.0
evaluation/subset.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"vhd": {"youtube_highlights": [231, 344, 236, 353, 73, 174, 277, 13, 187, 31, 182, 148, 172, 111, 11, 304, 167, 143, 146, 350, 276, 158, 107, 343, 99, 32], "qvhighlights": [486, 218, 117, 9, 392, 463, 1, 246, 2, 370, 406, 279, 264, 44, 409, 343, 213, 393, 497, 95, 64, 56, 72, 125, 170]}, "dvc": {"youcook2": [206, 119, 83, 303, 176, 318, 171, 223, 414, 183, 9, 362, 18, 234, 375, 396, 297, 41, 129, 402, 321, 158, 138, 130, 356, 189, 152, 53, 212], "hirest": [34, 28, 51, 45, 44, 33, 23, 8, 74, 47, 37, 32]}, "tem": {"qvhighlights": [129, 410, 107, 81, 196, 253, 306, 192, 139, 52, 322, 213, 183, 295, 226, 246, 184, 304, 19, 426, 282, 170, 408, 64, 347, 73, 76, 422, 201, 380, 61, 113], "perception_test": [65, 89, 312, 454, 256, 48, 463, 91, 476, 437, 9, 393, 281, 418, 254, 122, 98, 377, 471, 302, 263, 307, 482, 419, 343, 120, 395, 311, 361, 201, 38, 196]}, "eca": {"charades_sta": [210, 445, 53, 41, 386, 326, 255, 171, 173, 75, 102, 410, 486, 205, 95, 186, 157, 452, 304, 98, 136, 448]}, "tvg": {"charades_sta": [418, 361, 391, 294, 235, 68, 438, 195, 230, 247, 229, 300, 42, 301, 237, 225, 15, 403, 234, 489, 163, 56, 269, 472, 373, 316, 152, 356, 317, 103, 7, 54], "qvhighlights": [8, 349, 109, 37, 365, 58, 288, 116, 407, 378, 276, 431, 367, 44, 272, 359, 475, 388, 238, 188, 84, 14, 158, 124, 352, 251, 117, 27]}, "gvq": {"qa_ego4d": [286, 61, 33, 253, 72, 68, 260, 108, 262, 93, 236, 138, 200, 8, 139, 6, 119, 240, 289, 124, 27, 87, 206, 10]}, "epm": {"ego4d_nlq": [268, 490, 179, 104, 344, 150, 464, 39, 105, 431, 57, 4, 23, 117, 37, 347, 443, 447, 358, 470, 7, 119, 486, 131, 88, 495, 296, 234]}, "rvq": {"star": [465, 68, 137, 321, 381, 256, 38, 461, 43, 152, 317, 411, 487, 195, 76, 427, 312, 369, 345, 323, 211, 118, 51, 313, 75, 171]}, "slc": {"ht_step": [77, 239, 374, 240, 243, 206, 116, 316, 370, 106, 79, 194, 248, 312, 180, 211, 232, 252], "cross_task": [84, 36, 205, 253, 193, 144, 203, 260, 271, 126, 51, 155, 149, 5, 241, 146, 178, 95, 109, 61, 80, 33, 224, 274]}, "tal": {"thumos14": [94, 11, 75, 117, 86, 110, 120, 52, 131, 16, 116, 93, 46, 42, 9, 55, 122, 15, 78, 49, 61, 51, 123, 66, 129, 141, 72, 81], "thumos15": [129, 183, 32, 238, 262, 146, 6, 80, 253, 172, 221, 43, 139, 196, 9, 218, 50, 37, 132, 152, 260, 216, 95, 243, 237, 120, 247, 38, 256, 141], "perception_test": [392, 117, 251, 184, 24, 432, 5, 347, 262, 15, 170, 14, 81, 59, 438, 249, 417, 146, 336, 376, 287, 378, 386, 128, 125, 260, 334, 38]}, "rar": {"perception_test": [390, 184, 293, 352, 433, 48, 409, 402, 57, 404, 0, 308, 27, 46, 329, 126, 216, 171]}, "evs": {"tvsum": [11, 36, 6, 7, 30, 43, 38, 10]}}
videos/charades_sta.tar.gz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4fa4096b54479a2d5a75c60c7d1db8d19a6bdfbe77a08b5927e52ec21574c379
3
+ size 4972410880
videos/cross_task.tar.gz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c352e90bbe1f436049b6a8da675832a10b6eb01b368886dcc21457350de2c15
3
+ size 4513003520
videos/ego4d_nlq.tar.gz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:588ed1345b9aec523505045c44a893e81fd07cb76a85d4bf799693c41bfb3115
3
+ size 31775959040
videos/hirest.tar.gz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:510443570650a1051e4c3dfc4872b515c2960613d6af0d2eff9b7fc51705755d
3
+ size 1689518080