polinaeterna HF staff NimaBoscarino commited on
Commit
e19db67
0 Parent(s):

Duplicate from society-ethics/lila_camera_traps

Browse files

Co-authored-by: Nima Boscarino <[email protected]>

.gitattributes ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ *.7z filter=lfs diff=lfs merge=lfs -text
2
+ *.arrow filter=lfs diff=lfs merge=lfs -text
3
+ *.bin filter=lfs diff=lfs merge=lfs -text
4
+ *.bz2 filter=lfs diff=lfs merge=lfs -text
5
+ *.ftz filter=lfs diff=lfs merge=lfs -text
6
+ *.gz filter=lfs diff=lfs merge=lfs -text
7
+ *.h5 filter=lfs diff=lfs merge=lfs -text
8
+ *.joblib filter=lfs diff=lfs merge=lfs -text
9
+ *.lfs.* filter=lfs diff=lfs merge=lfs -text
10
+ *.lz4 filter=lfs diff=lfs merge=lfs -text
11
+ *.mlmodel filter=lfs diff=lfs merge=lfs -text
12
+ *.model filter=lfs diff=lfs merge=lfs -text
13
+ *.msgpack filter=lfs diff=lfs merge=lfs -text
14
+ *.npy filter=lfs diff=lfs merge=lfs -text
15
+ *.npz filter=lfs diff=lfs merge=lfs -text
16
+ *.onnx filter=lfs diff=lfs merge=lfs -text
17
+ *.ot filter=lfs diff=lfs merge=lfs -text
18
+ *.parquet filter=lfs diff=lfs merge=lfs -text
19
+ *.pb filter=lfs diff=lfs merge=lfs -text
20
+ *.pickle filter=lfs diff=lfs merge=lfs -text
21
+ *.pkl filter=lfs diff=lfs merge=lfs -text
22
+ *.pt filter=lfs diff=lfs merge=lfs -text
23
+ *.pth filter=lfs diff=lfs merge=lfs -text
24
+ *.rar filter=lfs diff=lfs merge=lfs -text
25
+ *.safetensors filter=lfs diff=lfs merge=lfs -text
26
+ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
27
+ *.tar.* filter=lfs diff=lfs merge=lfs -text
28
+ *.tflite filter=lfs diff=lfs merge=lfs -text
29
+ *.tgz filter=lfs diff=lfs merge=lfs -text
30
+ *.wasm filter=lfs diff=lfs merge=lfs -text
31
+ *.xz filter=lfs diff=lfs merge=lfs -text
32
+ *.zip filter=lfs diff=lfs merge=lfs -text
33
+ *.zst filter=lfs diff=lfs merge=lfs -text
34
+ *tfevents* filter=lfs diff=lfs merge=lfs -text
35
+ # Audio files - uncompressed
36
+ *.pcm filter=lfs diff=lfs merge=lfs -text
37
+ *.sam filter=lfs diff=lfs merge=lfs -text
38
+ *.raw filter=lfs diff=lfs merge=lfs -text
39
+ # Audio files - compressed
40
+ *.aac filter=lfs diff=lfs merge=lfs -text
41
+ *.flac filter=lfs diff=lfs merge=lfs -text
42
+ *.mp3 filter=lfs diff=lfs merge=lfs -text
43
+ *.ogg filter=lfs diff=lfs merge=lfs -text
44
+ *.wav filter=lfs diff=lfs merge=lfs -text
45
+ # Image files - uncompressed
46
+ *.bmp filter=lfs diff=lfs merge=lfs -text
47
+ *.gif filter=lfs diff=lfs merge=lfs -text
48
+ *.png filter=lfs diff=lfs merge=lfs -text
49
+ *.tiff filter=lfs diff=lfs merge=lfs -text
50
+ # Image files - compressed
51
+ *.jpg filter=lfs diff=lfs merge=lfs -text
52
+ *.jpeg filter=lfs diff=lfs merge=lfs -text
53
+ *.webp filter=lfs diff=lfs merge=lfs -text
54
+ data/Caltech_Camera_Traps.jsonl filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,541 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ annotations_creators:
3
+ - expert-generated
4
+ license:
5
+ - other
6
+ language_creators:
7
+ - expert-generated
8
+ language:
9
+ - en
10
+ multilinguality:
11
+ - monolingual
12
+ size_categories:
13
+ - 10M<n<100M
14
+ source_datasets:
15
+ - original
16
+ task_categories:
17
+ - image-classification
18
+ tags:
19
+ - biodiversity
20
+ - camera trap data
21
+ - wildlife monitoring
22
+ pretty_name: LILA Camera Traps
23
+ duplicated_from: society-ethics/lila_camera_traps
24
+ ---
25
+
26
+ # Dataset Card for LILA
27
+
28
+ ## Table of Contents
29
+ - [Table of Contents](#table-of-contents)
30
+ - [Dataset Description](#dataset-description)
31
+ - [Dataset Summary](#dataset-summary)
32
+ - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
33
+ - [Languages](#languages)
34
+ - [Dataset Structure](#dataset-structure)
35
+ - [Data Instances](#data-instances)
36
+ - [Data Fields](#data-fields)
37
+ - [Data Splits](#data-splits)
38
+ - [Dataset Creation](#dataset-creation)
39
+ - [Curation Rationale](#curation-rationale)
40
+ - [Source Data](#source-data)
41
+ - [Annotations](#annotations)
42
+ - [Personal and Sensitive Information](#personal-and-sensitive-information)
43
+ - [Considerations for Using the Data](#considerations-for-using-the-data)
44
+ - [Social Impact of Dataset](#social-impact-of-dataset)
45
+ - [Discussion of Biases](#discussion-of-biases)
46
+ - [Other Known Limitations](#other-known-limitations)
47
+ - [Additional Information](#additional-information)
48
+ - [Usage](#dataset-usage)
49
+ - [Dataset Curators](#dataset-curators)
50
+ - [Licensing Information](#licensing-information)
51
+ - [Citation Information](#citation-information)
52
+ - [Contributions](#contributions)
53
+
54
+ ## Dataset Description
55
+
56
+ - **Homepage:** https://lila.science/
57
+ - **Repository:** N/A
58
+ - **Paper:** N/A
59
+ - **Leaderboard:** N/A
60
+ - **Point of Contact:** [[email protected]]([email protected])
61
+
62
+ ### Dataset Summary
63
+
64
+ LILA Camera Traps is an aggregate data set of images taken by camera traps, which are devices that automatically (e.g. via motion detection) capture images of wild animals to help ecological research.
65
+
66
+ This data set is the first time when disparate camera trap data sets have been aggregated into a single training environment with a single [taxonomy](https://lila.science/taxonomy-mapping-for-camera-trap-data-sets/).
67
+
68
+ This data set consists of only camera trap image data sets, whereas the broader [LILA](lila.science/) website also has other data sets related to biology and conservation, intended as a resource for both machine learning (ML) researchers and those that want to harness ML for this topic.
69
+
70
+
71
+ See below for information about each specific dataset that LILA contains:
72
+
73
+ <details>
74
+ <summary> Caltech Camera Traps </summary>
75
+
76
+ This data set contains 243,100 images from 140 camera locations in the Southwestern United States, with labels for 21 animal categories (plus empty), primarily at the species level (for example, the most common labels are opossum, raccoon, and coyote), and approximately 66,000 bounding box annotations. Approximately 70% of images are labeled as empty.
77
+ More information about this data set is available [here](https://beerys.github.io/CaltechCameraTraps/).
78
+
79
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
80
+
81
+ For questions about this data set, contact [email protected].
82
+
83
+ If you use this data set, please cite the associated manuscript:
84
+ ```bibtex
85
+ @inproceedings{DBLP:conf/eccv/BeeryHP18,
86
+ author = {Sara Beery and
87
+ Grant Van Horn and
88
+ Pietro Perona},
89
+ title = {Recognition in Terra Incognita},
90
+ booktitle = {Computer Vision - {ECCV} 2018 - 15th European Conference, Munich,
91
+ Germany, September 8-14, 2018, Proceedings, Part {XVI}},
92
+ pages = {472--489},
93
+ year = {2018},
94
+ crossref = {DBLP:conf/eccv/2018-16},
95
+ url = {https://doi.org/10.1007/978-3-030-01270-0\_28},
96
+ doi = {10.1007/978-3-030-01270-0\_28},
97
+ timestamp = {Mon, 08 Oct 2018 17:08:07 +0200},
98
+ biburl = {https://dblp.org/rec/bib/conf/eccv/BeeryHP18},
99
+ bibsource = {dblp computer science bibliography, https://dblp.org}
100
+ }
101
+ ```
102
+ </details>
103
+
104
+ <details>
105
+ <summary> ENA24 </summary>
106
+
107
+ This data set contains approximately 10,000 camera trap images representing 23 classes from Eastern North America, with bounding boxes on each image. The most common classes are “American Crow”, “American Black Bear”, and “Dog”.
108
+
109
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
110
+
111
+ Please cite this manuscript if you use this data set:
112
+ ```bibtex
113
+ @article{yousif2019dynamic,
114
+ title={Dynamic Programming Selection of Object Proposals for Sequence-Level Animal Species Classification in the Wild},
115
+ author={Yousif, Hayder and Kays, Roland and He, Zhihai},
116
+ journal={IEEE Transactions on Circuits and Systems for Video Technology},
117
+ year={2019},
118
+ publisher={IEEE}
119
+ }
120
+ ```
121
+ For questions about this data set, contact [Hayder Yousif]([email protected]).
122
+
123
+ </details>
124
+
125
+ <details>
126
+ <summary> Missouri Camera Traps </summary>
127
+
128
+ This data set contains approximately 25,000 camera trap images representing 20 species (for example, the most common labels are red deer, mouflon, and white-tailed deer). Images within each sequence share the same species label (even though the animal may not have been recorded in all the images in the sequence). Around 900 bounding boxes are included. These are very challenging sequences with highly cluttered and dynamic scenes. Spatial resolutions of the images vary from 1920 × 1080 to 2048 × 1536. Sequence lengths vary from 3 to more than 300 frames.
129
+
130
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
131
+
132
+ If you use this data set, please cite the associated manuscript:
133
+ ```bibtex
134
+ @article{zhang2016animal,
135
+ title={Animal detection from highly cluttered natural scenes using spatiotemporal object region proposals and patch verification},
136
+ author={Zhang, Zhi and He, Zhihai and Cao, Guitao and Cao, Wenming},
137
+ journal={IEEE Transactions on Multimedia},
138
+ volume={18},
139
+ number={10},
140
+ pages={2079--2092},
141
+ year={2016},
142
+ publisher={IEEE}
143
+ }
144
+ ```
145
+ For questions about this data set, contact [Hayder Yousif]([email protected]) and [Zhi Zhang]([email protected]).
146
+ </details>
147
+
148
+ <details>
149
+ <summary> North American Camera Trap Images (NACTI) </summary>
150
+
151
+ This data set contains 3.7M camera trap images from five locations across the United States, with labels for 28 animal categories, primarily at the species level (for example, the most common labels are cattle, boar, and red deer). Approximately 12% of images are labeled as empty. We have also added bounding box annotations to 8892 images (mostly vehicles and birds).
152
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
153
+
154
+ Please cite this manuscript if you use this data set:
155
+ ```bibtex
156
+ @article{tabak2019machine,
157
+ title={Machine learning to classify animal species in camera trap images: Applications in ecology},
158
+ author={Tabak, Michael A and Norouzzadeh, Mohammad S and Wolfson, David W and Sweeney, Steven J and VerCauteren, Kurt C and Snow, Nathan P and Halseth, Joseph M and Di Salvo, Paul A and Lewis, Jesse S and White, Michael D and others},
159
+ journal={Methods in Ecology and Evolution},
160
+ volume={10},
161
+ number={4},
162
+ pages={585--590},
163
+ year={2019},
164
+ publisher={Wiley Online Library}
165
+ }
166
+ ```
167
+
168
+ For questions about this data set, contact [[email protected]]([email protected]).
169
+
170
+ </details>
171
+
172
+ <details>
173
+ <summary> WCS Camera Traps </summary>
174
+
175
+ This data set contains approximately 1.4M camera trap images representing around 675 species from 12 countries, making it one of the most diverse camera trap data sets available publicly. Data were provided by the [Wildlife Conservation Society](https://www.wcs.org/). The most common classes are tayassu pecari (peccary), meleagris ocellata (ocellated turkey), and bos taurus (cattle). A complete list of classes and associated image counts is available here. Approximately 50% of images are empty. We have also added approximately 375,000 bounding box annotations to approximately 300,000 of those images, which come from sequences covering almost all locations.
176
+
177
+ Sequences are inferred from timestamps, so may not strictly represent bursts. Images were labeled at a combination of image and sequence level, so – as is the case with most camera trap data sets – empty images may be labeled as non-empty (if an animal was present in one frame of a sequence but not in others). Images containing humans are referred to in metadata, but are not included in the data files. You can find more information about the data set [on the LILA website](https://lila.science/datasets/wcscameratraps).
178
+
179
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
180
+ </details>
181
+
182
+ <details>
183
+ <summary> Wellington Camera Traps </summary>
184
+
185
+ This data set contains 270,450 images from 187 camera locations in Wellington, New Zealand. The cameras (Bushnell 119537, 119476, and 119436) recorded sequences of three images when triggered. Each sequence was labelled by citizen scientists and/or professional ecologists from Victoria University of Wellington into 17 classes: 15 animal categories (for example, the most common labels are bird, cat, and hedgehog), empty, and unclassifiable. Approximately 17% of images are labeled as empty. Images within each sequence share the same species label (even though the animal may not have been recorded in all three images).
186
+
187
+ If you use this data set, please cite the associated manuscript:
188
+ ```bibtex
189
+ @article{anton2018monitoring,
190
+ title={Monitoring the mammalian fauna of urban areas using remote cameras and citizen science},
191
+ author={Anton, Victor and Hartley, Stephen and Geldenhuis, Andre and Wittmer, Heiko U},
192
+ journal={Journal of Urban Ecology},
193
+ volume={4},
194
+ number={1},
195
+ pages={juy002},
196
+ year={2018},
197
+ publisher={Oxford University Press}
198
+ }
199
+ ```
200
+
201
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
202
+
203
+ For questions about this data set, contact [Victor Anton]([email protected]).
204
+ </details>
205
+
206
+ <details>
207
+ <summary> Island Conservation Camera Traps </summary>
208
+
209
+ This data set contains approximately 123,000 camera trap images from 123 camera locations from 7 islands in 6 countries. Data were provided by Island Conservation during projects conducted to prevent the extinction of threatened species on islands.
210
+
211
+ The most common classes are rabbit, rat, petrel, iguana, cat, goat, and pig, with both rat and cat represented between multiple island sites representing significantly different ecosystems (tropical forest, dry forest, and temperate forests). Additionally, this data set represents data from locations and ecosystems that, to our knowledge, are not well represented in publicly available datasets including >1,000 images each of iguanas, petrels, and shearwaters. A complete list of classes and associated image counts is available here. Approximately 60% of the images are empty. We have also included approximately 65,000 bounding box annotations for about 50,000 images.
212
+
213
+ In general cameras were dispersed across each project site to detect the presence of invasive vertebrate species that threaten native island species. Cameras were set to capture bursts of photos for each motion detection event (between three and eight photos) with a set delay between events (10 to 30 seconds) to minimize the number of photos. Images containing humans are referred to in metadata, but are not included in the data files.
214
+
215
+ For questions about this data set, contact [David Will]([email protected]) at Island Conservation.
216
+
217
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
218
+
219
+ The original data set included a “human” class label; for privacy reasons, we have removed those images from this version of the data set. Those labels are still present in the metadata. If those images are important to your work, contact us; in some cases it will be possible to release those images under an alternative license.
220
+ </details>
221
+
222
+ <details>
223
+ <summary> Channel Islands Camera Traps </summary>
224
+
225
+ This data set contains 246,529 camera trap images from 73 camera locations in the Channel Islands, California. All animals are annotated with bounding boxes. Data were provided by The Nature Conservancy. Animals are classified as rodent1 (82914), fox (48150), bird (11099), skunk (1071), or other (159). 114,949 images (47%) are empty. All images of rats were taken on islands already known to have rat populations.
226
+
227
+ If you use these data in a publication or report, please use the following citation:
228
+
229
+ The Nature Conservancy (2021): Channel Islands Camera Traps 1.0. The Nature Conservancy. Dataset.
230
+
231
+ For questions about this data set, contact [Nathaniel Rindlaub]([email protected]) at The Nature Conservancy.
232
+
233
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
234
+
235
+ The original data set included a “human” class label; for privacy reasons, we have removed those images from this version of the data set. Those labels are still present in the metadata.
236
+
237
+ </details>
238
+
239
+ <details>
240
+ <summary> Idaho Camera Traps </summary>
241
+
242
+ This data set contains approximately 1.5 million camera trap images from Idaho. Labels are provided for 62 categories, most of which are animal classes (“deer”, “elk”, and “cattle” are the most common animal classes), but labels also include some state indicators (e.g. “snow on lens”, “foggy lens”). Approximately 70.5% of images are labeled as empty. Annotations were assigned to image sequences, rather than individual images, so annotations are meaningful only at the sequence level.
243
+
244
+ The metadata contains references to images containing humans, but these have been removed from the dataset (along with images containing vehicles and domestic dogs).
245
+
246
+ Images were provided by the Idaho Department of Fish and Game. No representations or warranties are made regarding the data, including but not limited to warranties of non-infringement or fitness for a particular purpose. Some information shared under this agreement may not have undergone quality assurance procedures and should be considered provisional. Images may not be sold in any format, but may be used for scientific publications. Please acknowledge the Idaho Department of Fish and Game when using images for publication or scientific communication.
247
+ </details>
248
+
249
+ <details>
250
+ <summary> Snapshot Serengeti </summary>
251
+
252
+ This data set contains approximately 2.65M sequences of camera trap images, totaling 7.1M images, from seasons one through eleven of the [Snapshot Serengeti project](https://snapshotserengeti.org/) -- the flagship project of the Snapshot Safari network. Using the same camera trapping protocols at every site, Snapshot Safari members are collecting standardized data from many protected areas in Africa, which allows for cross-site comparisons to assess the efficacy of conservation and restoration programs. Serengeti National Park in Tanzania is best known for the massive annual migrations of wildebeest and zebra that drive the cycling of its dynamic ecosystem.
253
+
254
+ Labels are provided for 61 categories, primarily at the species level (for example, the most common labels are wildebeest, zebra, and Thomson’s gazelle). Approximately 76% of images are labeled as empty. A full list of species and associated image counts is available [here](https://lilablobssc.blob.core.windows.net/snapshotserengeti-v-2-0/SnapshotSerengeti_S1-11_v2.1.species_list.csv). We have also added approximately 150,000 bounding box annotations to approximately 78,000 of those images.
255
+
256
+ The images and species-level labels are described in more detail in the associated manuscript:
257
+
258
+ ```bibtex
259
+ @misc{dryad_5pt92,
260
+ title = {Data from: Snapshot Serengeti, high-frequency annotated camera trap images of 40 mammalian species in an African savanna},
261
+ author = {Swanson, AB and Kosmala, M and Lintott, CJ and Simpson, RJ and Smith, A and Packer, C},
262
+ year = {2015},
263
+ journal = {Scientific Data},
264
+ URL = {https://doi.org/10.5061/dryad.5pt92},
265
+ doi = {doi:10.5061/dryad.5pt92},
266
+ publisher = {Dryad Digital Repository}
267
+ }
268
+ ```
269
+
270
+ For questions about this data set, contact [Sarah Huebner]([email protected]) at the University of Minnesota.
271
+
272
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
273
+ </details>
274
+
275
+ <details>
276
+ <summary> Snapshot Karoo </summary>
277
+
278
+ This data set contains 14889 sequences of camera trap images, totaling 38074 images, from the [Snapshot Karoo](https://www.zooniverse.org/projects/shuebner729/snapshot-karoo) project, part of the Snapshot Safari network. Using the same camera trapping protocols at every site, Snapshot Safari members are collecting standardized data from many protected areas in Africa, which allows for cross-site comparisons to assess the efficacy of conservation and restoration programs. Karoo National Park, located in the arid Nama Karoo biome of South Africa, is defined by its endemic vegetation and mountain landscapes. Its unique topographical gradient has led to a surprising amount of biodiversity, with 58 mammals and more than 200 bird species recorded, as well as a multitude of reptilian species.
279
+
280
+ Labels are provided for 38 categories, primarily at the species level (for example, the most common labels are gemsbokoryx, hartebeestred, and kudu). Approximately 83.02% of images are labeled as empty. A full list of species and associated image counts is available [here](https://lilablobssc.blob.core.windows.net/snapshot-safari/KAR/SnapshotKaroo_S1_v1.0.species_list.csv).
281
+
282
+ For questions about this data set, contact [Sarah Huebner]([email protected]) at the University of Minnesota.
283
+
284
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
285
+ </details>
286
+
287
+
288
+ <details>
289
+ <summary> Snapshot Kgalagadi </summary>
290
+
291
+ This data set contains 3611 sequences of camera trap images, totaling 10222 images, from the [Snapshot Kgalagadi](https://www.zooniverse.org/projects/shuebner729/snapshot-kgalagadi/) project, part of the Snapshot Safari network. Using the same camera trapping protocols at every site, Snapshot Safari members are collecting standardized data from many protected areas in Africa, which allows for cross-site comparisons to assess the efficacy of conservation and restoration programs. The Kgalagadi Transfrontier Park stretches from the Namibian border across South Africa and into Botswana, covering a landscape commonly referred to as the Kalahari – an arid savanna. This region is of great interest to help us understand how animals cope with extreme temperatures at both ends of the scale.
292
+
293
+ Labels are provided for 31 categories, primarily at the species level (for example, the most common labels are gemsbokoryx, birdother, and ostrich). Approximately 76.14% of images are labeled as empty. A full list of species and associated image counts is available [here](https://lilablobssc.blob.core.windows.net/snapshot-safari/KGA/SnapshotKgalagadi_S1_v1.0.species_list.csv).
294
+
295
+ For questions about this data set, contact [Sarah Huebner]([email protected]) at the University of Minnesota.
296
+
297
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
298
+ </details>
299
+
300
+
301
+ <details>
302
+ <summary> Snapshot Enonkishu </summary>
303
+
304
+ This data set contains 13301 sequences of camera trap images, totaling 28544 images, from the [Snapshot Enonkishu](https://www.zooniverse.org/projects/aguthmann/snapshot-enonkishu) project, part of the Snapshot Safari network. Using the same camera trapping protocols at every site, Snapshot Safari members are collecting standardized data from many protected areas in Africa, which allows for cross-site comparisons to assess the efficacy of conservation and restoration programs. Enonkishu Conservancy is located on the northern boundary of the Mara-Serengeti ecosystem in Kenya, and is managed by a consortium of stakeholders and land-owning Maasai families. Their aim is to promote coexistence between wildlife and livestock in order to encourage regenerative grazing and build stability in the Mara conservancies.
305
+
306
+ Labels are provided for 39 categories, primarily at the species level (for example, the most common labels are impala, warthog, and zebra). Approximately 64.76% of images are labeled as empty. A full list of species and associated image counts is available [here](https://lilablobssc.blob.core.windows.net/snapshot-safari/ENO/SnapshotEnonkishu_S1_v1.0.species_list.csv).
307
+
308
+ For questions about this data set, contact [Sarah Huebner]([email protected]) at the University of Minnesota.
309
+
310
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
311
+ </details>
312
+
313
+
314
+ <details>
315
+ <summary> Snapshot Camdeboo </summary>
316
+
317
+ This data set contains 12132 sequences of camera trap images, totaling 30227 images, from the [Snapshot Camdeboo](https://www.zooniverse.org/projects/shuebner729/snapshot-camdeboo) project, part of the Snapshot Safari network. Using the same camera trapping protocols at every site, Snapshot Safari members are collecting standardized data from many protected areas in Africa, which allows for cross-site comparisons to assess the efficacy of conservation and restoration programs. Camdeboo National Park, South Africa is crucial habitat for many birds on a global scale, with greater than fifty endemic and near-endemic species and many migratory species.
318
+
319
+ Labels are provided for 43 categories, primarily at the species level (for example, the most common labels are kudu, springbok, and ostrich). Approximately 43.74% of images are labeled as empty. A full list of species and associated image counts is available [here](https://lilablobssc.blob.core.windows.net/snapshot-safari/CDB/SnapshotCamdeboo_S1_v1.0.species_list.csv).
320
+
321
+ For questions about this data set, contact [Sarah Huebner]([email protected]) at the University of Minnesota.
322
+
323
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
324
+ </details>
325
+
326
+
327
+ <details>
328
+ <summary> Snapshot Mountain Zebra </summary>
329
+
330
+ This data set contains 71688 sequences of camera trap images, totaling 73034 images, from the [Snapshot Mountain Zebra](https://www.zooniverse.org/projects/meredithspalmer/snapshot-mountain-zebra/) project, part of the Snapshot Safari network. Using the same camera trapping protocols at every site, Snapshot Safari members are collecting standardized data from many protected areas in Africa, which allows for cross-site comparisons to assess the efficacy of conservation and restoration programs. Mountain Zebra National Park is located in the Eastern Cape of South Africa in a transitional area between several distinct biomes, which means it is home to many endemic species. As the name suggests, this park contains the largest remnant population of Cape Mountain zebras, ~700 as of 2019 and increasing steadily every year.
331
+
332
+ Labels are provided for 54 categories, primarily at the species level (for example, the most common labels are zebramountain, kudu, and springbok). Approximately 91.23% of images are labeled as empty. A full list of species and associated image counts is available [here](https://lilablobssc.blob.core.windows.net/snapshot-safari/MTZ/SnapshotMountainZebra_S1_v1.0.species_list.csv).
333
+
334
+ For questions about this data set, contact [Sarah Huebner]([email protected]) at the University of Minnesota.
335
+
336
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
337
+ </details>
338
+
339
+
340
+ <details>
341
+ <summary> Snapshot Kruger </summary>
342
+
343
+ This data set contains 4747 sequences of camera trap images, totaling 10072 images, from the [Snapshot Kruger](https://www.zooniverse.org/projects/shuebner729/snapshot-kruger) project, part of the Snapshot Safari network. Using the same camera trapping protocols at every site, Snapshot Safari members are collecting standardized data from many protected areas in Africa, which allows for cross-site comparisons to assess the efficacy of conservation and restoration programs. Kruger National Park, South Africa has been a refuge for wildlife since its establishment in 1898, and it houses one of the most diverse wildlife assemblages remaining in Africa. The Snapshot Safari grid was established in 2018 as part of a research project assessing the impacts of large mammals on plant life as boundary fences were removed and wildlife reoccupied areas of previous extirpation.
344
+
345
+ Labels are provided for 46 categories, primarily at the species level (for example, the most common labels are impala, elephant, and buffalo). Approximately 61.60% of images are labeled as empty. A full list of species and associated image counts is available [here](https://lilablobssc.blob.core.windows.net/snapshot-safari/KRU/SnapshotKruger_S1_v1.0.species_list.csv).
346
+
347
+ For questions about this data set, contact [Sarah Huebner]([email protected]) at the University of Minnesota.
348
+
349
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
350
+ </details>
351
+
352
+
353
+ <details>
354
+ <summary> SWG Camera Traps </summary>
355
+
356
+ This data set contains 436,617 sequences of camera trap images from 982 locations in Vietnam and Lao, totaling 2,039,657 images. Labels are provided for 120 categories, primarily at the species level (for example, the most common labels are “Eurasian Wild Pig”, “Large-antlered Muntjac”, and “Unidentified Murid”). Approximately 12.98% of images are labeled as empty. A full list of species and associated image counts is available here. 101,659 bounding boxes are provided on 88,135 images.
357
+
358
+ This data set is provided by the Saola Working Group; providers include:
359
+
360
+ - IUCN SSC Asian Wild Cattle Specialist Group’s Saola Working Group (SWG)
361
+ - Asian Arks
362
+ - Wildlife Conservation Society (Lao)
363
+ - WWF Lao
364
+ - Integrated Conservation of Biodiversity and Forests project, Lao (ICBF)
365
+ - Center for Environment and Rural Development, Vinh University, Vietnam
366
+
367
+ If you use these data in a publication or report, please use the following citation:
368
+
369
+ SWG (2021): Northern and Central Annamites Camera Traps 2.0. IUCN SSC Asian Wild Cattle Specialist Group’s Saola Working Group. Dataset.
370
+
371
+ For questions about this data set, contact [email protected].
372
+
373
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
374
+
375
+ </details>
376
+
377
+ <details>
378
+ <summary> Orinoquia Camera Traps </summary>
379
+
380
+ This data set contains 104,782 images collected from a 50-camera-trap array deployed from January to July 2020 within the private natural reserves El Rey Zamuro (31 km2) and Las Unamas (40 km2), located in the Meta department in the Orinoquía region in central Colombia. We deployed cameras using a stratified random sampling design across forest core area strata. Cameras were spaced 1 km apart from one another, located facing wildlife trails, and deployed with no bait. Images were stored and reviewed by experts using the Wildlife Insights platform.
381
+
382
+ This data set contains 51 classes, predominantly mammals such as the collared peccary, black agouti, spotted paca, white-lipped peccary, lowland tapir, and giant anteater. Approximately 20% of images are empty.
383
+
384
+ The main purpose of the study is to understand how humans, wildlife, and domestic animals interact in multi-functional landscapes (e.g., agricultural livestock areas with native forest remnants). However, this data set was also used to review model performance of AI-powered platforms – Wildlife Insights (WI), MegaDetector (MD), and Machine Learning for Wildlife Image Classification (MLWIC2). We provide a demonstration of the use of WI, MD, and MLWIC2 and R code for evaluating model performance of these platforms in the accompanying [GitHub repository](https://github.com/julianavelez1/Processing-Camera-Trap-Data-Using-AI).
385
+
386
+ If you use these data in a publication or report, please use the following citation:
387
+ ```bibtex
388
+ @article{velez2022choosing,
389
+ title={Choosing an Appropriate Platform and Workflow for Processing Camera Trap Data using Artificial Intelligence},
390
+ author={V{\'e}lez, Juliana and Castiblanco-Camacho, Paula J and Tabak, Michael A and Chalmers, Carl and Fergus, Paul and Fieberg, John},
391
+ journal={arXiv preprint arXiv:2202.02283},
392
+ year={2022}
393
+ }
394
+ ```
395
+ For questions about this data set, contact [Juliana Velez Gomez]([email protected]).
396
+
397
+ This data set is released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/).
398
+ </details>
399
+
400
+ ### Supported Tasks and Leaderboards
401
+
402
+ No leaderboards exist for LILA.
403
+
404
+ ### Languages
405
+
406
+ The [LILA taxonomy](https://lila.science/taxonomy-mapping-for-camera-trap-data-sets/) is provided in English.
407
+
408
+ ## Dataset Structure
409
+
410
+ ### Data Instances
411
+
412
+ The data annotations are provided in [COCO Camera Traps](https://github.com/Microsoft/CameraTraps/blob/master/data_management/README.md#coco-cameratraps-format) format.
413
+
414
+ All of the datasets share a common category taxonomy, which is defined on the [LILA website](https://lila.science/taxonomy-mapping-for-camera-trap-data-sets/).
415
+
416
+ ### Data Fields
417
+
418
+ Different datasets may have slightly varying fields, which include:
419
+
420
+ `file_name`: the file name \
421
+ `width` and `height`: the dimensions of the image \
422
+ `study`: which research study the image was collected as part of \
423
+ `location` : the name of the location at which the image was taken \
424
+ `annotations`: information about image annotation, which includes the taxonomy information, bounding box/boxes (`bbox`/`bboxes`) if any, as well as any other annotation information. \
425
+ `image` : the `path` to download the image and any other information that is available, e.g. its size in `bytes`.
426
+
427
+
428
+ ### Data Splits
429
+
430
+ This dataset does not have a predefined train/test split.
431
+
432
+ ## Dataset Creation
433
+
434
+ ### Curation Rationale
435
+
436
+ The datasets that constitute LILA have been provided by the organizations, projects and researchers who collected them.
437
+
438
+ ### Source Data
439
+
440
+ #### Initial data collection and normalization
441
+
442
+ N/A
443
+
444
+ #### Who are the source language producers?
445
+
446
+ N/A
447
+
448
+ ### Annotations
449
+
450
+ #### Annotation process
451
+
452
+ Each dataset has been annotated by the members of the project/organization that provided it.
453
+
454
+ #### Who are the annotators?
455
+
456
+ The annotations have been provided by domain experts in fields such as biology and ecology.
457
+
458
+ ### Personal and Sensitive Information
459
+
460
+ Some of the original data sets included a “human” class label; for privacy reasons, these images were removed. Those labels are still present in the metadata. If those images are important to your work, contact the [LILA maintainers](mailto:[email protected]), since in some cases it will be possible to release those images under an alternative license.
461
+
462
+ ## Considerations for Using the Data
463
+
464
+ ### Social Impact of Dataset
465
+
466
+ Machine learning depends on labeled data, but accessing such data in biology and conservation is a challenge. Consequently, everyone benefits when labeled data is made available. Biologists and conservation scientists benefit by having data to train on, and free hosting allows teams to multiply the impact of their data (we suggest listing this benefit in grant proposals that fund data collection). ML researchers benefit by having data to experiment with.
467
+
468
+ ### Discussion of Biases
469
+
470
+ These datasets do not represent global diversity, but are examples of local ecosystems and animals.
471
+
472
+ ### Other Known Limitations
473
+
474
+ N/A
475
+
476
+ ## Additional Information
477
+
478
+ ### Working with Taxonomies
479
+
480
+ All the taxonomy categories are saved as ClassLabels, which can be converted to strings as needed. Strings can likewise be converted to integers as needed, to filter the dataset. In the example below we filter the "Caltech Camera Traps" dataset to find all the entries with a "felis catus" as the species for the first annotation.
481
+
482
+ ```python
483
+ dataset = load_dataset("society-ethics/lila_camera_traps", "Caltech Camera Traps", split="train")
484
+ taxonomy = dataset.features["annotations"].feature["taxonomy"]
485
+
486
+ # Filters to show only cats
487
+ cats = dataset.filter(lambda x: x["annotations"]["taxonomy"][0]["species"] == taxonomy["species"].str2int("felis catus"))
488
+ ```
489
+
490
+ The original common names have been saved with their taxonomy mappings in this repository in `common_names_to_tax.json`. These can be used, for example, to map from a taxonomy combination to a common name to help make queries more legible. Note, however, that there is a small number of duplicate common names with different taxonomy values which you will need to disambiguate.
491
+
492
+ The following example loads the first "sea turtle" in the "Island Conservation Camera Traps" dataset.
493
+
494
+ ```python
495
+ LILA_COMMON_NAMES_TO_TAXONOMY = pd.read_json("https://huggingface.co/datasets/society-ethics/lila_camera_traps/raw/main/data/common_names_to_tax.json", lines=True).set_index("common_name")
496
+ dataset = load_dataset("society-ethics/lila_camera_traps", "Island Conservation Camera Traps", split="train")
497
+ taxonomy = dataset.features["annotations"].feature["taxonomy"]
498
+
499
+ sea_turtle = LILA_COMMON_NAMES_TO_TAXONOMY.loc["sea turtle"].to_dict()
500
+ sea_turtle = {k: taxonomy[k].str2int(v) if v is not None else v for k, v in sea_turtle.items()} # Map to ClassLabel integers
501
+
502
+ sea_turtle_dataset = ds.filter(lambda x: x["annotations"]["taxonomy"][0] == sea_turtle)
503
+ ```
504
+
505
+ The example below selects a random item from the dataset, and then maps from the taxonomy to a common name:
506
+
507
+ ```python
508
+ LILA_COMMON_NAMES_TO_TAXONOMY = pd.read_json("https://huggingface.co/datasets/society-ethics/lila_camera_traps/raw/main/data/common_names_to_tax.json", lines=True).set_index("common_name")
509
+
510
+ dataset = load_dataset("society-ethics/lila_camera_traps", "Caltech Camera Traps", split="train")
511
+ taxonomy = dataset.features["annotations"].feature["taxonomy"]
512
+
513
+ random_entry = dataset.shuffle()[0]
514
+ filter_taxonomy = random_entry["annotations"]["taxonomy"][0]
515
+
516
+ filter_keys = list(map(lambda x: (x[0], taxonomy[x[0]].int2str(x[1])), filter(lambda x: x[1] is not None, list(filter_taxonomy.items()))))
517
+
518
+ if len(filter_keys) > 0:
519
+ print(LILA_COMMON_NAMES_TO_TAXONOMY[np.logical_and.reduce([
520
+ LILA_COMMON_NAMES_TO_TAXONOMY[k] == v for k,v in filter_keys
521
+ ])])
522
+ else:
523
+ print("No common name found for the item.")
524
+ ```
525
+
526
+
527
+ ### Dataset Curators
528
+
529
+ LILA BC is maintained by a working group that includes representatives from Ecologize, Zooniverse, the Evolving AI Lab, Snapshot Safari, and Microsoft AI for Earth. Hosting on Microsoft Azure is provided by Microsoft AI for Earth.
530
+
531
+ ### Licensing Information
532
+
533
+ Many, but not all, LILA data sets were released under the [Community Data License Agreement (permissive variant)](https://cdla.io/permissive-1-0/). Check the details of the specific dataset you are using in its section above.
534
+
535
+ ### Citation Information
536
+
537
+ Citations for each dataset (if they exist) are provided in its section above.
538
+
539
+ ### Contributions
540
+
541
+ Thanks to [@NimaBoscarino](https://github.com/NimaBoscarino/) for adding this dataset.
data/Caltech_Camera_Traps.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7719462753aa532feff4dcad77cfafb62e30f2b81a079836639c9ea46900ebdd
3
+ size 10247472
data/Channel_Islands_Camera_Traps.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d178c7ce8af9dd90025af7066c1e31134c115e53e2f231414a32518affdf02de
3
+ size 6248572
data/ENA24.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ed833705bfd7078678ea46ab83bb26feb4e9cbd286ee4d8cf79f85a5c163f91
3
+ size 334216
data/Idaho_Camera_Traps.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7cfb3c74eb5eae3747f7344aa965270af2fd65d10173174af69b969425fad39e
3
+ size 18310491
data/Island_Conservation_Camera_Traps.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fa313a38af48d6c48fc7a90b9c9126c8cb329b9ae0ad89707f6d0bd71b5ab72d
3
+ size 2461480
data/Missouri_Camera_Traps.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:70e9f8e411df98b5f177670cfd8a2136a12e5cf758a2ccc1c593e18d31772a6f
3
+ size 256710
data/NACTI.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:069d804624122b416dfec24d98d5cc2607272898601ee3a5b56bfde5181670e7
3
+ size 18865808
data/Orinoquia_Camera_Traps.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:79d55d7daf639b6e70bf091bc99f1976c1694d09575fe96c2bca7169b0280ae8
3
+ size 1923112
data/README.md ADDED
File without changes
data/SWG_Camera_Traps.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:317c9f3da7ea066607d110093ac69c55e9c72fd31490c788d93b8bd7176c31a3
3
+ size 36197933
data/Snapshot_Camdeboo.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7f4f064bcc95b40c1acf184ae14719128a4ab5eb5471f3ec411143ef27e603f6
3
+ size 509493
data/Snapshot_Enonkishu.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:18d75056f810c95de0fc67da79440009881ce6e01bf1ba4c42a886a39245762e
3
+ size 586186
data/Snapshot_Karoo.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7e887ec06a76f9425512b820532bcadf4cd28db85ebbdb1e1820dfe57648d9c5
3
+ size 629130
data/Snapshot_Kgalagadi.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9dcc6310db1191defff771292cb6724094d9e258faa5f58259ab2b6559ca8118
3
+ size 177282
data/Snapshot_Kruger.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:87bda0de2458003ed8d3fde35ee6ce483d4d74ca28b1aedf0100d57a709f1bac
3
+ size 217010
data/Snapshot_Mountain_Zebra.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e8fca90878f588e032d279fb329975aed864d70d1ba4aa12173c16ab26efd50
3
+ size 2018000
data/Snapshot_Serengeti.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:baa676037c1bad598284b39fa83b949a4b34f73ac1da65d4e040821c238e9d23
3
+ size 127297784
data/WCS_Camera_Traps.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef57f7ccabc0b8a94c37f05824cabfff95480fc97fa96f586219f7525d6b70fb
3
+ size 24699486
data/Wellington_Camera_Traps.jsonl.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f24e5732f8342cd4f9aab5692a24b960e678dfd0e53d33a1623aab2681558f7b
3
+ size 6169258
data/common_names_to_tax.json ADDED
The diff for this file is too large to render. See raw diff
 
lila_camera_traps.py ADDED
@@ -0,0 +1,929 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
2
+ #
3
+ # Licensed under the Apache License, Version 2.0 (the "License");
4
+ # you may not use this file except in compliance with the License.
5
+ # You may obtain a copy of the License at
6
+ #
7
+ # http://www.apache.org/licenses/LICENSE-2.0
8
+ #
9
+ # Unless required by applicable law or agreed to in writing, software
10
+ # distributed under the License is distributed on an "AS IS" BASIS,
11
+ # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
12
+ # See the License for the specific language governing permissions and
13
+ # limitations under the License.
14
+ """
15
+ LILA Camera Traps is an aggregate data set of images taken by camera traps, which are devices that automatically (e.g. via motion detection) capture images of wild animals to help ecological research.
16
+
17
+ This data set is the first time when disparate camera trap data sets have been aggregated into a single training environment with a single taxonomy.
18
+
19
+ This data set consists of only camera trap image data sets, whereas the broader LILA website (https://lila.science) also has other data sets related to biology and conservation, intended as a resource for both machine learning (ML) researchers and those that want to harness ML for this topic.
20
+ """
21
+
22
+ import json
23
+ import os
24
+ import pandas as pd
25
+
26
+ import datasets
27
+
28
+ _LILA_CITATIONS = {
29
+ "Caltech Camera Traps": """
30
+ @inproceedings{DBLP:conf/eccv/BeeryHP18,
31
+ author = {Sara Beery and
32
+ Grant Van Horn and
33
+ Pietro Perona},
34
+ title = {Recognition in Terra Incognita},
35
+ booktitle = {Computer Vision - {ECCV} 2018 - 15th European Conference, Munich,
36
+ Germany, September 8-14, 2018, Proceedings, Part {XVI}},
37
+ pages = {472--489},
38
+ year = {2018},
39
+ crossref = {DBLP:conf/eccv/2018-16},
40
+ url = {https://doi.org/10.1007/978-3-030-01270-0\_28},
41
+ doi = {10.1007/978-3-030-01270-0\_28},
42
+ timestamp = {Mon, 08 Oct 2018 17:08:07 +0200},
43
+ biburl = {https://dblp.org/rec/bib/conf/eccv/BeeryHP18},
44
+ bibsource = {dblp computer science bibliography, https://dblp.org}
45
+ }
46
+ """,
47
+ "ENA24": """
48
+ @article{yousif2019dynamic,
49
+ title={Dynamic Programming Selection of Object Proposals for Sequence-Level Animal Species Classification in the Wild},
50
+ author={Yousif, Hayder and Kays, Roland and He, Zhihai},
51
+ journal={IEEE Transactions on Circuits and Systems for Video Technology},
52
+ year={2019},
53
+ publisher={IEEE}
54
+ }
55
+ """,
56
+ "Missouri Camera Traps": """
57
+ @article{zhang2016animal,
58
+ title={Animal detection from highly cluttered natural scenes using spatiotemporal object region proposals and patch verification},
59
+ author={Zhang, Zhi and He, Zhihai and Cao, Guitao and Cao, Wenming},
60
+ journal={IEEE Transactions on Multimedia},
61
+ volume={18},
62
+ number={10},
63
+ pages={2079--2092},
64
+ year={2016},
65
+ publisher={IEEE}
66
+ }
67
+ """,
68
+ "NACTI": """
69
+ @article{tabak2019machine,
70
+ title={Machine learning to classify animal species in camera trap images: Applications in ecology},
71
+ author={Tabak, Michael A and Norouzzadeh, Mohammad S and Wolfson, David W and Sweeney, Steven J and VerCauteren, Kurt C and Snow, Nathan P and Halseth, Joseph M and Di Salvo, Paul A and Lewis, Jesse S and White, Michael D and others},
72
+ journal={Methods in Ecology and Evolution},
73
+ volume={10},
74
+ number={4},
75
+ pages={585--590},
76
+ year={2019},
77
+ publisher={Wiley Online Library}
78
+ }
79
+ """,
80
+ "WCS Camera Traps": "",
81
+ "Wellington Camera Traps": """
82
+ @article{anton2018monitoring,
83
+ title={Monitoring the mammalian fauna of urban areas using remote cameras and citizen science},
84
+ author={Anton, Victor and Hartley, Stephen and Geldenhuis, Andre and Wittmer, Heiko U},
85
+ journal={Journal of Urban Ecology},
86
+ volume={4},
87
+ number={1},
88
+ pages={juy002},
89
+ year={2018},
90
+ publisher={Oxford University Press}
91
+ }
92
+ """,
93
+ "Island Conservation Camera Traps": "",
94
+ "Channel Islands Camera Traps": "",
95
+ "Idaho Camera Traps": "",
96
+ "Snapshot Serengeti": """
97
+ @misc{dryad_5pt92,
98
+ title = {Data from: Snapshot Serengeti, high-frequency annotated camera trap images of 40 mammalian species in an African savanna},
99
+ author = {Swanson, AB and Kosmala, M and Lintott, CJ and Simpson, RJ and Smith, A and Packer, C},
100
+ year = {2015},
101
+ journal = {Scientific Data},
102
+ URL = {https://doi.org/10.5061/dryad.5pt92},
103
+ doi = {doi:10.5061/dryad.5pt92},
104
+ publisher = {Dryad Digital Repository}
105
+ }
106
+ """,
107
+ "Snapshot Karoo": "",
108
+ "Snapshot Kgalagadi": "",
109
+ "Snapshot Enonkishu": "",
110
+ "Snapshot Camdeboo": "",
111
+ "Snapshot Mountain Zebra": "",
112
+ "Snapshot Kruger": "",
113
+ "SWG Camera Traps": "",
114
+ "Orinoquia Camera Traps": """
115
+ @article{velez2022choosing,
116
+ title={Choosing an Appropriate Platform and Workflow for Processing Camera Trap Data using Artificial Intelligence},
117
+ author={V{\'e}lez, Juliana and Castiblanco-Camacho, Paula J and Tabak, Michael A and Chalmers, Carl and Fergus, Paul and Fieberg, John},
118
+ journal={arXiv preprint arXiv:2202.02283},
119
+ year={2022}
120
+ }
121
+ """,
122
+ }
123
+
124
+ # You can copy an official description
125
+ _DESCRIPTION = """\
126
+ LILA Camera Traps is an aggregate data set of images taken by camera traps, which are devices that automatically (e.g. via motion detection) capture images of wild animals to help ecological research.
127
+
128
+ This data set is the first time when disparate camera trap data sets have been aggregated into a single training environment with a single taxonomy.
129
+
130
+ This data set consists of only camera trap image data sets, whereas the broader LILA website also has other data sets related to biology and conservation, intended as a resource for both machine learning (ML) researchers and those that want to harness ML for this topic.
131
+ """
132
+
133
+ _HOMEPAGE = "https://huggingface.co/datasets/society-ethics/LILA"
134
+
135
+ _LILA_SAS_URLS = pd.read_csv("https://lila.science/wp-content/uploads/2020/03/lila_sas_urls.txt")
136
+ _LILA_SAS_URLS.rename(columns={"# name": "name"}, inplace=True)
137
+
138
+ _METADATA_BASE_URL = "https://huggingface.co/datasets/NimaBoscarino/LILA/resolve/main/data/"
139
+
140
+ _LILA_URLS = {
141
+ "Caltech Camera Traps": "Caltech_Camera_Traps.jsonl.zip",
142
+ "ENA24": "ENA24.jsonl.zip",
143
+ "Missouri Camera Traps": "Missouri_Camera_Traps.jsonl.zip",
144
+ "NACTI": "NACTI.jsonl.zip",
145
+ "WCS Camera Traps": "WCS_Camera_Traps.jsonl.zip",
146
+ "Wellington Camera Traps": "Wellington_Camera_Traps.jsonl.zip",
147
+ "Island Conservation Camera Traps": "Island_Conservation_Camera_Traps.jsonl.zip",
148
+ "Channel Islands Camera Traps": "Channel_Islands_Camera_Traps.jsonl.zip",
149
+ "Idaho Camera Traps": "Idaho_Camera_Traps.jsonl.zip",
150
+ "Snapshot Serengeti": "Snapshot_Serengeti.jsonl.zip",
151
+ "Snapshot Karoo": "Snapshot_Karoo.jsonl.zip",
152
+ "Snapshot Kgalagadi": "Snapshot_Kgalagadi.jsonl.zip",
153
+ "Snapshot Enonkishu": "Snapshot_Enonkishu.jsonl.zip",
154
+ "Snapshot Camdeboo": "Snapshot_Camdeboo.jsonl.zip",
155
+ "Snapshot Mountain Zebra": "Snapshot_Mountain_Zebra.jsonl.zip",
156
+ "Snapshot Kruger": "Snapshot_Kruger.jsonl.zip",
157
+ "SWG Camera Traps": "SWG_Camera_Traps.jsonl.zip",
158
+ "Orinoquia Camera Traps": "Orinoquia_Camera_Traps.jsonl.zip",
159
+ }
160
+
161
+ _TAXONOMY = {
162
+ "kingdom": datasets.ClassLabel(num_classes=1, names=["animalia"]),
163
+ "phylum": datasets.ClassLabel(num_classes=2, names=["chordata", "arthropoda"]),
164
+ "subphylum": datasets.ClassLabel(num_classes=5, names=[
165
+ 'vertebrata', 'hexapoda', 'crustacea', 'chelicerata',
166
+ 'myriapoda'
167
+ ]),
168
+ "superclass": datasets.ClassLabel(num_classes=1, names=["multicrustacea"]),
169
+ "class": datasets.ClassLabel(num_classes=8, names=[
170
+ 'mammalia', 'aves', 'insecta', 'reptilia', 'malacostraca',
171
+ 'arachnida', 'diplopoda', 'amphibia'
172
+ ]),
173
+ "subclass": datasets.ClassLabel(num_classes=3, names=[
174
+ 'theria', 'pterygota', 'eumalacostraca'
175
+ ]),
176
+ "infraclass": datasets.ClassLabel(num_classes=2, names=[
177
+ 'placentalia', 'marsupialia'
178
+ ]),
179
+ "superorder": datasets.ClassLabel(num_classes=5, names=[
180
+ 'laurasiatheria', 'euarchontoglires', 'eucarida', 'xenarthra',
181
+ 'afrotheria'
182
+ ]),
183
+ "order": datasets.ClassLabel(num_classes=53, names=[
184
+ 'carnivora', 'chiroptera', 'artiodactyla', 'squamata',
185
+ 'didelphimorphia', 'lagomorpha', 'rodentia', 'primates',
186
+ 'passeriformes', 'galliformes', 'perissodactyla',
187
+ 'accipitriformes', 'caprimulgiformes', 'lepidoptera',
188
+ 'strigiformes', 'piciformes', 'falconiformes', 'charadriiformes',
189
+ 'decapoda', 'columbiformes', 'pelecaniformes', 'procellariiformes',
190
+ 'gruiformes', 'testudines', 'araneae', 'tinamiformes', 'cingulata',
191
+ 'coraciiformes', 'hymenoptera', 'pilosa', 'cathartiformes',
192
+ 'tubulidentata', 'otidiformes', 'struthioniformes', 'proboscidea',
193
+ 'crocodylia', 'pholidota', 'scandentia', 'trogoniformes',
194
+ 'bucerotiformes', 'anseriformes', 'eulipotyphla', 'psittaciformes',
195
+ 'cuculiformes', 'ciconiiformes', 'musophagiformes', 'hyracoidea',
196
+ 'eurypygiformes', 'afrosoricida', 'galbuliformes', 'macroscelidea',
197
+ 'anura', 'rheiformes'
198
+ ]),
199
+ "suborder": datasets.ClassLabel(num_classes=17, names=[
200
+ 'ruminantia', 'suina', 'sciuromorpha', 'haplorhini',
201
+ 'hystricomorpha', 'pleocyemata', 'sauria', 'myomorpha',
202
+ 'castorimorpha', 'apocrita', 'vermilingua', 'anomaluromorpha',
203
+ 'whippomorpha', 'serpentes', 'tylopoda', 'strepsirrhini',
204
+ 'tenrecomorpha'
205
+ ]),
206
+ "infraorder": datasets.ClassLabel(num_classes=9, names=[
207
+ 'simiiformes', 'hystricognathi', 'brachyura', 'anomura',
208
+ 'aculeata', 'ancodonta', 'chiromyiformes', 'lemuriformes',
209
+ 'lorisiformes'
210
+ ]),
211
+ "superfamily": datasets.ClassLabel(num_classes=12, names=[
212
+ 'hominoidea', 'erethizontoidea', 'paguroidea', 'muroidea',
213
+ 'chelonioidea', 'cavioidea', 'formicoidea', 'octodontoidea',
214
+ 'lemuroidea', 'chinchilloidea', 'cheirogaleoidea', 'papilionoidea'
215
+ ]),
216
+ "family": datasets.ClassLabel(num_classes=159, names=[
217
+ 'mustelidae', 'felidae', 'bovidae', 'canidae', 'cervidae',
218
+ 'didelphidae', 'suidae', 'leporidae', 'procyonidae', 'mephitidae',
219
+ 'sciuridae', 'hominidae', 'ursidae', 'corvidae', 'phasianidae',
220
+ 'equidae', 'turdidae', 'accipitridae', 'trochilidae',
221
+ 'erethizontidae', 'antilocapridae', 'sittidae', 'parulidae',
222
+ 'cardinalidae', 'picidae', 'falconidae', 'strigidae', 'laridae',
223
+ 'columbidae', 'ardeidae', 'calcinidae', 'iguanidae',
224
+ 'megapodiidae', 'mimidae', 'varanidae', 'procellariidae',
225
+ 'rallidae', 'muridae', 'phocidae', 'hydrobatidae', 'dasyproctidae',
226
+ 'tayassuidae', 'tinamidae', 'cuniculidae', 'odontophoridae',
227
+ 'dasypodidae', 'passerellidae', 'troglodytidae', 'cricetidae',
228
+ 'geomyidae', 'momotidae', 'formicidae', 'caviidae', 'cracidae',
229
+ 'myrmecophagidae', 'chlamyphoridae', 'tapiridae', 'cebidae',
230
+ 'pitheciidae', 'cathartidae', 'atelidae', 'caprimulgidae',
231
+ 'orycteropodidae', 'hyaenidae', 'cercopithecidae', 'otididae',
232
+ 'gruidae', 'viverridae', 'pedetidae', 'herpestidae',
233
+ 'struthionidae', 'hystricidae', 'sagittariidae', 'testudinidae',
234
+ 'elephantidae', 'giraffidae', 'hippopotamidae', 'rhinocerotidae',
235
+ 'crocodylidae', 'numididae', 'manidae', 'irenidae', 'echimyidae',
236
+ 'pittidae', 'leiothrichidae', 'muscicapidae', 'tragulidae',
237
+ 'scolopacidae', 'hylobatidae', 'timaliidae', 'stenostiridae',
238
+ 'tupaiidae', 'trogonidae', 'bucerotidae', 'prionodontidae',
239
+ 'acrocephalidae', 'pycnonotidae', 'anatidae', 'anhimidae',
240
+ 'anomaluridae', 'aramidae', 'erinaceidae', 'brachypteraciidae',
241
+ 'threskiornithidae', 'psittacidae', 'buphagidae', 'burhinidae',
242
+ 'camelidae', 'sarothruridae', 'cuculidae', 'ciconiidae',
243
+ 'furnariidae', 'cisticolidae', 'apodidae', 'musophagidae',
244
+ 'nesomyidae', 'eupleridae', 'daubentoniidae', 'procaviidae',
245
+ 'dicaeidae', 'dicruridae', 'lemuridae', 'laniidae', 'vangidae',
246
+ 'eurypygidae', 'formicariidae', 'galagidae', 'grallariidae',
247
+ 'charadriidae', 'tenrecidae', 'scotocercidae', 'chinchillidae',
248
+ 'sturnidae', 'malaconotidae', 'macrosphenidae', 'cheirogaleidae',
249
+ 'alaudidae', 'icteridae', 'bucconidae', 'motacillidae',
250
+ 'nandiniidae', 'nectariniidae', 'estrildidae', 'bernieridae',
251
+ 'alligatoridae', 'macroscelididae', 'ploceidae', 'indriidae',
252
+ 'psophiidae', 'ramphastidae', 'ranidae', 'rheidae', 'spalacidae',
253
+ 'scincidae', 'soricidae', 'monarchidae', 'thryonomyidae',
254
+ 'teiidae', 'tytonidae'
255
+ ]),
256
+ "subfamily": datasets.ClassLabel(num_classes=69, names=[
257
+ 'taxidiinae', 'felinae', 'bovinae', 'capreolinae',
258
+ 'didelphinae', 'suinae', 'sciurinae', 'homininae', 'ursinae',
259
+ 'xerinae', 'mephitinae', 'antilopinae', 'cervinae', 'mustelinae',
260
+ 'guloninae', 'erethizontinae', 'sterninae', 'ardeinae', 'murinae',
261
+ 'lutrinae', 'melinae', 'neotominae', 'hydrochoerinae',
262
+ 'tigriornithinae', 'tolypeutinae', 'pantherinae', 'cebinae',
263
+ 'callicebinae', 'alouattinae', 'saimiriinae', 'protelinae',
264
+ 'cercopithecinae', 'genettinae', 'mungotinae', 'herpestinae',
265
+ 'ictonychinae', 'hyaeninae', 'mellivorinae', 'echimyinae',
266
+ 'paradoxurinae', 'ratufinae', 'helictidinae', 'colobinae',
267
+ 'viverrinae', 'hemigalinae', 'callosciurinae', 'erinaceinae',
268
+ 'atelinae', 'camelinae', 'caviinae', 'furnariinae', 'criniferinae',
269
+ 'cricetomyinae', 'euplerinae', 'deomyinae', 'nesomyinae',
270
+ 'euphractinae', 'galidiinae', 'tenrecinae', 'oryzorictinae',
271
+ 'musophaginae', 'myadinae', 'macroscelidinae', 'rhizomyinae',
272
+ 'rhynchocyoninae', 'scincinae', 'crocidurinae', 'tremarctinae',
273
+ 'tupinambinae'
274
+ ]),
275
+ "tribe": datasets.ClassLabel(num_classes=46, names=[
276
+ 'bovini', 'odocoileini', 'didelphini', 'suini', 'sciurini',
277
+ 'tamiini', 'marmotini', 'caprini', 'cervini', 'alceini', 'rattini',
278
+ 'capreolini', 'apodemini', 'reithrodontomyini', 'neotomini',
279
+ 'papionini', 'alcelaphini', 'potamochoerini', 'cephalophini',
280
+ 'tragelaphini', 'hippotragini', 'oreotragini', 'cercopithecini',
281
+ 'reduncini', 'antilopini', 'aepycerotini', 'phacochoerini',
282
+ 'xerini', 'echimyini', 'pteromyini', 'presbytini', 'muntiacini',
283
+ 'callosciurini', 'camelini', 'colobini', 'praomyini',
284
+ 'protoxerini', 'arvicanthini', 'malacomyini', 'metachirini',
285
+ 'murini', 'neotragini', 'macroscelidini', 'myocastorini',
286
+ 'rhizomyini', 'lamini'
287
+ ]),
288
+ "genus": datasets.ClassLabel(num_classes=476, names=[
289
+ 'taxidea', 'lynx', 'felis', 'bos', 'canis', 'odocoileus',
290
+ 'urocyon', 'puma', 'didelphis', 'sus', 'procyon', 'sciurus',
291
+ 'homo', 'ursus', 'corvus', 'gallus', 'tamias', 'sylvilagus',
292
+ 'equus', 'vulpes', 'mephitis', 'meleagris', 'marmota', 'ovis',
293
+ 'sialia', 'nucifraga', 'cervus', 'mustela', 'pekania', 'neogale',
294
+ 'pica', 'alces', 'erethizon', 'antilocapra', 'sitta', 'ixoreus',
295
+ 'piranga', 'falco', 'strix', 'anous', 'athene', 'nasua', 'capra',
296
+ 'ardea', 'butorides', 'calcinus', 'iguana', 'caloenas', 'rattus',
297
+ 'calonectris', 'asio', 'hydrobates', 'zenaida', 'nyctanassa',
298
+ 'turdus', 'dasyprocta', 'pecari', 'lepus', 'tinamus', 'leopardus',
299
+ 'cuniculus', 'mazama', 'tamiasciurus', 'capreolus', 'apodemus',
300
+ 'callipepla', 'cyanocitta', 'dasypus', 'dendragapus', 'junco',
301
+ 'lontra', 'martes', 'meles', 'otospermophilus', 'perisoreus',
302
+ 'troglodytes', 'peromyscus', 'neotoma', 'momotus', 'speothos',
303
+ 'hydrochoerus', 'cerdocyon', 'mitu', 'tigrisoma', 'myrmecophaga',
304
+ 'priodontes', 'pteronura', 'panthera', 'herpailurus', 'tapirus',
305
+ 'sapajus', 'plecturocebus', 'tamandua', 'penelope', 'eira',
306
+ 'cathartes', 'alouatta', 'saimiri', 'tayassu', 'orycteropus',
307
+ 'proteles', 'papio', 'damaliscus', 'syncerus', 'potamochoerus',
308
+ 'ardeotis', 'caracal', 'anthropoides', 'sylvicapra', 'tragelaphus',
309
+ 'dama', 'otocyon', 'oryx', 'genetta', 'pedetes', 'alcelaphus',
310
+ 'lupulella', 'oreotragus', 'suricata', 'herpestes', 'cynictis',
311
+ 'chlorocebus', 'struthio', 'hystrix', 'redunca', 'pelea',
312
+ 'sagittarius', 'antidorcas', 'raphicerus', 'connochaetes',
313
+ 'ictonyx', 'acinonyx', 'madoqua', 'cephalophus', 'loxodonta',
314
+ 'nanger', 'eudorcas', 'giraffa', 'hippopotamus', 'crocuta',
315
+ 'aepyceros', 'ourebia', 'phacochoerus', 'kobus', 'neotis',
316
+ 'parahyaena', 'bunolagus', 'diceros', 'mellivora', 'crocodylus',
317
+ 'pronolagus', 'hippotragus', 'leptailurus', 'lycaon', 'xerus',
318
+ 'ceratotherium', 'hyaena', 'nesolagus', 'irena', 'atherurus',
319
+ 'macaca', 'dactylomys', 'hydrornis', 'macropygia', 'varanus',
320
+ 'arctictis', 'ratufa', 'pterorhinus', 'cinclidium', 'myophonus',
321
+ 'moschiola', 'capricornis', 'cissa', 'paradoxurus', 'urva',
322
+ 'rheinardia', 'spilornis', 'chalcophaps', 'scolopax', 'melogale',
323
+ 'enicurus', 'trachypithecus', 'petaurista', 'cyanoderma',
324
+ 'catopuma', 'garrulax', 'culicicapa', 'polyplectron', 'arctonyx',
325
+ 'muntiacus', 'viverra', 'erythrogenys', 'prionailurus', 'picus',
326
+ 'pardofelis', 'paguma', 'nisaetus', 'ducula', 'tupaia',
327
+ 'harpactes', 'geokichla', 'chrotogale', 'callosciurus', 'manis',
328
+ 'dremomys', 'pygathrix', 'trochalopteron', 'ianthocincla',
329
+ 'aceros', 'rusa', 'zoothera', 'leiothrix', 'lophura', 'prionodon',
330
+ 'helarctos', 'pitta', 'tamiops', 'myiomela', 'urocissa',
331
+ 'accipiter', 'acrocephalus', 'acryllium', 'agamia', 'alectoris',
332
+ 'chamaetylas', 'alophoixus', 'alopochen', 'stelgidillas',
333
+ 'eurillas', 'anhima', 'anomalurus', 'aonyx', 'aquila', 'aramides',
334
+ 'aramus', 'arborophila', 'arctogalidia', 'ardeola', 'argusianus',
335
+ 'arremonops', 'atelerix', 'ateles', 'atelocynus', 'atelornis',
336
+ 'atilax', 'balearica', 'bambusicola', 'baryphthengus', 'bdeogale',
337
+ 'blastocerus', 'bostrychia', 'brachypteracias', 'brotogeris',
338
+ 'bubo', 'bubulcus', 'buphagus', 'burhinus', 'butastur', 'buteo',
339
+ 'buteogallus', 'bycanistes', 'cabassous', 'cairina', 'caloperdix',
340
+ 'camelus', 'mentocrex', 'caprimulgus', 'caracara', 'carpococcyx',
341
+ 'hylocichla', 'catharus', 'cavia', 'cebus', 'cercocebus',
342
+ 'cercopithecus', 'allochrocebus', 'cercotrichas', 'ortalis',
343
+ 'chelonoidis', 'ciconia', 'cinclodes', 'circus', 'cisticola',
344
+ 'civettictis', 'claravis', 'cochlearius', 'coendou', 'collocalia',
345
+ 'colobus', 'colomys', 'columba', 'columbina', 'conepatus',
346
+ 'copsychus', 'coragyps', 'corythaixoides', 'cossypha', 'coturnix',
347
+ 'coua', 'crax', 'cricetomys', 'cryptoprocta', 'crypturellus',
348
+ 'cuon', 'cyanoptila', 'cyornis', 'daptrius', 'daubentonia',
349
+ 'dendrocitta', 'dendrohyrax', 'ortygornis', 'deomys', 'dicaeum',
350
+ 'dicerorhinus', 'dicrurus', 'melaenornis', 'egretta', 'elephas',
351
+ 'eliurus', 'larvivora', 'erythrocebus', 'eulemur', 'euphractus',
352
+ 'eupleres', 'eupodotis', 'eurocephalus', 'euryceros', 'eurypyga',
353
+ 'eutriorchis', 'ficedula', 'formicarius', 'fossa', 'scleroptila',
354
+ 'pternistis', 'francolinus', 'funisciurus', 'galago', 'galictis',
355
+ 'galidia', 'galidictis', 'geotrygon', 'grallaria', 'guttera',
356
+ 'haliaeetus', 'vanellus', 'harpia', 'heliosciurus', 'helogale',
357
+ 'hemicentetes', 'hemigalus', 'urosphena', 'heterohyrax',
358
+ 'hippocamelus', 'hybomys', 'hylomyscus', 'hylopetes', 'hypogeomys',
359
+ 'ichneumia', 'arundinax', 'jynx', 'lagidium', 'lamprotornis',
360
+ 'laniarius', 'lanius', 'lariscus', 'lemur', 'leptotila',
361
+ 'lissotis', 'litocranius', 'lophotibis', 'lutreolina', 'lycalopex',
362
+ 'malacomys', 'melierax', 'melocichla', 'mesembrinibis',
363
+ 'chloropicus', 'metachirus', 'micrastur', 'microcebus',
364
+ 'microgale', 'microsciurus', 'mirafra', 'molothrus', 'monasa',
365
+ 'morphnus', 'motacilla', 'mungos', 'mus', 'musophaga', 'mydaus',
366
+ 'myoprocta', 'mystacornis', 'nandinia', 'cyanomitra', 'oressochen',
367
+ 'neocossyphus', 'neofelis', 'neomorphus', 'delacourella',
368
+ 'streptopelia', 'nesomys', 'nesotragus', 'niltava', 'nothocrax',
369
+ 'numida', 'nyctidromus', 'odontophorus', 'oenomys', 'oenanthe',
370
+ 'otolemur', 'otus', 'oxylabes', 'paleosuchus', 'pan', 'paraxerus',
371
+ 'pernis', 'petrodromus', 'phaethornis', 'philander', 'philantomba',
372
+ 'pilherodius', 'xanthomixis', 'pipile', 'ploceus', 'poecilogale',
373
+ 'pogonocichla', 'potos', 'praomys', 'presbytis', 'procavia',
374
+ 'piliocolobus', 'proechimys', 'propithecus', 'protoxerus',
375
+ 'psophia', 'pteroglossus', 'ramphastos', 'rana', 'rhea',
376
+ 'rhizomys', 'rhynchocyon', 'rollulus', 'rupornis', 'ruwenzorornis',
377
+ 'salanoia', 'saxicola', 'setifer', 'sheppardia', 'plestiodon',
378
+ 'spilogale', 'spizaetus', 'stephanoaetus', 'stigmochelys',
379
+ 'amazona', 'suncus', 'sundasciurus', 'tauraco', 'tenrec',
380
+ 'terpsiphone', 'thamnomys', 'thryonomys', 'tockus', 'tolypeutes',
381
+ 'tragulus', 'tremarctos', 'trichys', 'tupinambis', 'turtur',
382
+ 'tyto', 'vicugna', 'viverricula', 'xenoperdix', 'euxerus',
383
+ 'zonotrichia', 'erinaceus'
384
+ ]),
385
+ "species": datasets.ClassLabel(num_classes=668, names=[
386
+ 'taxidea taxus', 'lynx rufus', 'felis catus', 'bos taurus',
387
+ 'canis latrans', 'canis familiaris', 'urocyon cinereoargenteus',
388
+ 'puma concolor', 'didelphis virginiana', 'sus scrofa',
389
+ 'procyon lotor', 'urocyon littoralis', 'homo sapiens',
390
+ 'ursus americanus', 'corvus brachyrhynchos', 'gallus gallus',
391
+ 'tamias striatus', 'sylvilagus floridanus', 'sciurus niger',
392
+ 'sciurus carolinensis', 'equus caballus', 'vulpes vulpes',
393
+ 'mephitis mephitis', 'odocoileus virginianus',
394
+ 'meleagris gallopavo', 'marmota monax', 'ovis canadensis',
395
+ 'nucifraga columbiana', 'cervus canadensis', 'mustela erminea',
396
+ 'pekania pennanti', 'neogale frenata', 'pica hudsonia',
397
+ 'alces alces', 'erethizon dorsatum', 'antilocapra americana',
398
+ 'corvus corax', 'sitta canadensis', 'ixoreus naevius',
399
+ 'piranga ludoviciana', 'canis lupus', 'falco sparverius',
400
+ 'strix varia', 'anous stolidus', 'athene cunicularia',
401
+ 'nasua nasua', 'equus asinus', 'capra hircus', 'ardea herodias',
402
+ 'butorides virescens', 'calcinus tubularis', 'falco tinnunculus',
403
+ 'caloenas nicobarica', 'asio flammeus', 'hydrobates pelagicus',
404
+ 'zenaida asiatica', 'nyctanassa violacea', 'dasyprocta coibae',
405
+ 'pecari tajacu', 'didelphis marsupialis', 'lepus europaeus',
406
+ 'tinamus major', 'ovis ammon', 'leopardus pardalis',
407
+ 'mazama americana', 'cervus elaphus', 'tamiasciurus hudsonicus',
408
+ 'rattus praetor', 'nasua narica', 'apodemus sylvaticus',
409
+ 'callipepla californica', 'cyanocitta stelleri',
410
+ 'dasypus novemcinctus', 'dendragapus obscurus', 'equus africanus',
411
+ 'equus ferus', 'junco hyemalis', 'lepus americanus',
412
+ 'lepus californicus', 'lontra canadensis', 'marmota flaviventris',
413
+ 'martes americana', 'meles meles', 'odocoileus hemionus',
414
+ 'otospermophilus beecheyi', 'perisoreus canadensis',
415
+ 'rattus rattus', 'troglodytes aedon', 'zenaida macroura',
416
+ 'momotus momota', 'dasyprocta fuliginosa', 'speothos venaticus',
417
+ 'hydrochoerus hydrochaeris', 'iguana iguana', 'cerdocyon thous',
418
+ 'mitu tomentosum', 'tigrisoma fasciatum',
419
+ 'myrmecophaga tridactyla', 'priodontes maximus',
420
+ 'pteronura brasiliensis', 'panthera onca',
421
+ 'herpailurus yagouaroundi', 'tapirus terrestris', 'sapajus apella',
422
+ 'leopardus wiedii', 'lontra longicaudis', 'sciurus igniventris',
423
+ 'dasyprocta guamara', 'plecturocebus ornatus', 'mitu salvini',
424
+ 'tamandua tetradactyla', 'penelope jacquacu', 'cuniculus paca',
425
+ 'eira barbara', 'cathartes aura', 'penelope jacucaca',
426
+ 'tayassu pecari', 'orycteropus afer', 'proteles cristatus',
427
+ 'damaliscus pygargus', 'syncerus caffer', 'potamochoerus larvatus',
428
+ 'ardeotis kori', 'caracal caracal', 'anthropoides paradiseus',
429
+ 'sylvicapra grimmia', 'tragelaphus oryx', 'dama dama',
430
+ 'otocyon megalotis', 'oryx gazella', 'lepus saxatilis',
431
+ 'pedetes capensis', 'alcelaphus buselaphus', 'lupulella mesomelas',
432
+ 'oreotragus oreotragus', 'tragelaphus strepsiceros',
433
+ 'suricata suricatta', 'herpestes ichneumon',
434
+ 'cynictis penicillata', 'chlorocebus pygerythrus',
435
+ 'struthio camelus', 'hystrix africaeaustralis',
436
+ 'redunca fulvorufula', 'pelea capreolus',
437
+ 'sagittarius serpentarius', 'antidorcas marsupialis',
438
+ 'raphicerus campestris', 'connochaetes gnou', 'equus zebra',
439
+ 'ictonyx striatus', 'tragelaphus scriptus', 'acinonyx jubatus',
440
+ 'loxodonta africana', 'nanger granti', 'eudorcas thomsonii',
441
+ 'giraffa camelopardalis', 'lepus victoriae',
442
+ 'hippopotamus amphibius', 'crocuta crocuta', 'aepyceros melampus',
443
+ 'panthera pardus', 'panthera leo', 'ourebia ourebi',
444
+ 'hystrix cristata', 'damaliscus lunatus', 'phacochoerus africanus',
445
+ 'kobus ellipsiprymnus', 'connochaetes taurinus', 'equus quagga',
446
+ 'neotis ludwigii', 'vulpes chama', 'parahyaena brunnea',
447
+ 'herpestes pulverulentus', 'bunolagus monticularis',
448
+ 'diceros bicornis', 'felis lybica', 'lepus capensis',
449
+ 'mellivora capensis', 'crocodylus niloticus',
450
+ 'cephalophus natalensis', 'lupulella adusta',
451
+ 'tragelaphus angasii', 'pronolagus randensis',
452
+ 'hippotragus equinus', 'leptailurus serval', 'lycaon pictus',
453
+ 'ceratotherium simum', 'hyaena hyaena', 'nesolagus timminsi',
454
+ 'irena puella', 'ursus thibetanus', 'atherurus macrourus',
455
+ 'mustela strigidorsa', 'hydrornis elliotii', 'macropygia unchall',
456
+ 'varanus bengalensis', 'arctictis binturong', 'ratufa bicolor',
457
+ 'pterorhinus chinensis', 'cinclidium frontale',
458
+ 'hydrornis cyaneus', 'myophonus caeruleus', 'strix leptogrammica',
459
+ 'moschiola meminna', 'capricornis sumatraensis', 'cissa chinensis',
460
+ 'paradoxurus hermaphroditus', 'urva urva', 'rheinardia ocellata',
461
+ 'spilornis cheela', 'chalcophaps indica', 'scolopax rusticola',
462
+ 'turdus obscurus', 'trachypithecus francoisi',
463
+ 'cyanoderma chrysaeum', 'catopuma temminckii', 'garrulax maesi',
464
+ 'culicicapa ceylonensis', 'polyplectron bicalcaratum',
465
+ 'trachypithecus hatinhensis', 'arctonyx collaris',
466
+ 'cissa hypoleuca', 'turdus cardis', 'muntiacus vuquangensis',
467
+ 'viverra zibetha', 'erythrogenys hypoleucos',
468
+ 'prionailurus bengalensis', 'picus chlorolophus',
469
+ 'hystrix brachyura', 'pardofelis marmorata', 'paguma larvata',
470
+ 'nisaetus nipalensis', 'ducula badia', 'pterorhinus pectoralis',
471
+ 'tupaia belangeri', 'harpactes oreskios', 'geokichla citrina',
472
+ 'chrotogale owstoni', 'callosciurus erythraeus',
473
+ 'trachypithecus phayrei', 'macaca nemestrina',
474
+ 'dremomys rufigenis', 'picus rabieri', 'muntiacus muntjak',
475
+ 'pygathrix nemaeus', 'trochalopteron milnei',
476
+ 'muntiacus rooseveltorum', 'garrulax castanotis',
477
+ 'ianthocincla konkakinhensis', 'aceros nipalensis',
478
+ 'rusa unicolor', 'zoothera dauma', 'geokichla sibirica',
479
+ 'leiothrix argentauris', 'lophura nycthemera',
480
+ 'prionodon pardicolor', 'butorides striata', 'macaca arctoides',
481
+ 'helarctos malayanus', 'enicurus leschenaulti', 'myiomela leucura',
482
+ 'urocissa whiteheadi', 'mustela kathiah', 'martes flavigula',
483
+ 'accipiter madagascariensis', 'accipiter melanoleucus',
484
+ 'acrocephalus baeticatus', 'acryllium vulturinum', 'agamia agami',
485
+ 'alectoris rufa', 'chamaetylas poliophrys', 'alophoixus bres',
486
+ 'alopochen aegyptiaca', 'alouatta sara',
487
+ 'stelgidillas gracilirostris', 'eurillas latirostris',
488
+ 'eurillas virens', 'anhima cornuta', 'anomalurus derbianus',
489
+ 'aonyx cinereus', 'aquila heliaca', 'aquila rapax',
490
+ 'aramides cajaneus', 'aramus guarauna',
491
+ 'arborophila brunneopectus', 'arborophila rubrirostris',
492
+ 'arborophila rufogularis', 'arctogalidia trivirgata',
493
+ 'arctonyx hoevenii', 'ardea alba', 'ardea cocoi',
494
+ 'ardea melanocephala', 'ardeola grayii', 'argusianus argus',
495
+ 'arremonops chloronotus', 'asio madagascariensis',
496
+ 'atelerix albiventris', 'ateles chamek', 'atelocynus microtis',
497
+ 'atelornis pittoides', 'atherurus africanus', 'atilax paludinosus',
498
+ 'balearica regulorum', 'bambusicola fytchii',
499
+ 'baryphthengus martii', 'bdeogale crassicauda',
500
+ 'bdeogale jacksoni', 'blastocerus dichotomus', 'bos gaurus',
501
+ 'bostrychia hagedash', 'brachypteracias squamiger',
502
+ 'bubulcus ibis', 'burhinus capensis', 'butastur indicus',
503
+ 'buteo ridgwayi', 'buteogallus urubitinga', 'bycanistes brevis',
504
+ 'cabassous centralis', 'cabassous unicinctus', 'cairina moschata',
505
+ 'callosciurus notatus', 'caloperdix oculeus',
506
+ 'camelus dromedarius', 'mentocrex kioloides', 'capra aegagrus',
507
+ 'caracara plancus', 'carpococcyx renauldi',
508
+ 'cathartes burrovianus', 'cathartes melambrotus',
509
+ 'hylocichla mustelina', 'catharus ustulatus', 'cavia aperea',
510
+ 'cebus albifrons', 'cephalophus harveyi', 'cephalophus nigrifrons',
511
+ 'cephalophus silvicultor', 'cephalophus spadix',
512
+ 'cercocebus sanjei', 'cercopithecus erythrogaster',
513
+ 'allochrocebus lhoesti', 'cercopithecus mitis', 'ortalis vetula',
514
+ 'chelonoidis carbonarius', 'ciconia maguari',
515
+ 'cinclodes atacamensis', 'cinclodes fuscus', 'circus cyaneus',
516
+ 'cisticola cherina', 'civettictis civetta', 'claravis pretiosa',
517
+ 'cochlearius cochlearius', 'coendou bicolor', 'collocalia linchi',
518
+ 'colobus angolensis', 'colomys goslingi', 'columba arquatrix',
519
+ 'columba larvata', 'columbina talpacoti', 'conepatus chinga',
520
+ 'conepatus semistriatus', 'copsychus albospecularis',
521
+ 'copsychus malabaricus', 'copsychus saularis', 'coragyps atratus',
522
+ 'corythaixoides leucogaster', 'cossypha archeri',
523
+ 'coturnix delegorguei', 'coua caerulea', 'coua ruficeps',
524
+ 'coua serriana', 'crax alector', 'crax rubra',
525
+ 'cricetomys gambianus', 'cryptoprocta ferox',
526
+ 'crypturellus atrocapillus', 'crypturellus boucardi',
527
+ 'crypturellus cinereus', 'crypturellus cinnamomeus',
528
+ 'crypturellus erythropus', 'crypturellus bartletti',
529
+ 'crypturellus soui', 'crypturellus undulatus',
530
+ 'crypturellus variegatus', 'cuniculus taczanowskii',
531
+ 'cuon alpinus', 'cyanoptila cyanomelana', 'cyornis banyumas',
532
+ 'daptrius ater', 'dasyprocta punctata', 'dasyprocta leporina',
533
+ 'dasypus kappleri', 'daubentonia madagascariensis',
534
+ 'dendrocitta occipitalis', 'dendrohyrax arboreus',
535
+ 'ortygornis sephaena', 'deomys ferrugineus',
536
+ 'dicaeum trigonostigma', 'dicerorhinus sumatrensis',
537
+ 'dicrurus adsimilis', 'didelphis imperfecta', 'didelphis pernigra',
538
+ 'melaenornis fischeri', 'egretta thula', 'elephas maximus',
539
+ 'eliurus penicillatus', 'eliurus petteri', 'eliurus webbi',
540
+ 'enicurus schistaceus', 'equus grevyi', 'larvivora cyane',
541
+ 'erythrocebus patas', 'eudorcas rufifrons', 'eulemur albifrons',
542
+ 'euphractus sexcinctus', 'eupleres goudotii',
543
+ 'eupodotis senegalensis', 'eurocephalus ruppelli',
544
+ 'euryceros prevostii', 'eurypyga helias', 'eutriorchis astur',
545
+ 'felis chaus', 'felis silvestris', 'ficedula mugimaki',
546
+ 'ficedula tricolor', 'formicarius analis', 'formicarius colma',
547
+ 'fossa fossana', 'scleroptila afra', 'pternistis nobilis',
548
+ 'funisciurus carruthersi', 'funisciurus pyrropus',
549
+ 'galago senegalensis', 'galictis vittata', 'galidia elegans',
550
+ 'galidictis fasciata', 'genetta genetta', 'genetta maculata',
551
+ 'genetta servalina', 'genetta tigrina', 'geokichla gurneyi',
552
+ 'geotrygon montana', 'geotrygon saphirina', 'grallaria andicolus',
553
+ 'guttera pucherani', 'haliaeetus vociferoides', 'vanellus cayanus',
554
+ 'harpia harpyja', 'buteogallus solitarius',
555
+ 'heliosciurus rufobrachium', 'heliosciurus ruwenzorii',
556
+ 'helogale parvula', 'hemicentetes semispinosus',
557
+ 'hemigalus derbyanus', 'urosphena neumanni',
558
+ 'herpestes sanguineus', 'urva semitorquata', 'heterohyrax brucei',
559
+ 'hippocamelus antisensis', 'hybomys univittatus',
560
+ 'hydrornis oatesi', 'hylomyscus stella', 'hylopetes alboniger',
561
+ 'hypogeomys antimena', 'ichneumia albicauda', 'arundinax aedon',
562
+ 'jynx torquilla', 'lagidium viscacia', 'lamprotornis chalybaeus',
563
+ 'lamprotornis hildebrandti', 'lamprotornis superbus',
564
+ 'laniarius funebris', 'lanius collaris', 'lariscus insignis',
565
+ 'leopardus tigrinus', 'leptotila plumbeiceps',
566
+ 'leptotila rufaxilla', 'leptotila verreauxi',
567
+ 'lissotis hartlaubii', 'lissotis melanogaster',
568
+ 'litocranius walleri', 'lophotibis cristata', 'eupodotis gindiana',
569
+ 'lophura diardi', 'lophura erythrophthalma', 'lophura ignita',
570
+ 'lophura inornata', 'lutreolina crassicaudata',
571
+ 'lycalopex culpaeus', 'macaca assamensis', 'macaca fascicularis',
572
+ 'macaca mulatta', 'madoqua guentheri', 'malacomys longipes',
573
+ 'manis javanica', 'mazama temama', 'mazama chunyi',
574
+ 'mazama gouazoubira', 'odocoileus pandora',
575
+ 'melaenornis ardesiacus', 'melaenornis pammelaina',
576
+ 'meleagris ocellata', 'melierax poliopterus',
577
+ 'melocichla mentalis', 'melogale everetti', 'melogale personata',
578
+ 'mesembrinibis cayennensis', 'chloropicus griseocephalus',
579
+ 'metachirus nudicaudatus', 'microcebus murinus',
580
+ 'microsciurus flaviventer', 'microsciurus mimulus',
581
+ 'mitu tuberosum', 'molothrus oryzivorus', 'monasa morphoeus',
582
+ 'morphnus guianensis', 'motacilla flava', 'motacilla flaviventris',
583
+ 'mungos mungo', 'mus minutoides', 'musophaga rossae',
584
+ 'mustela lutreolina', 'mydaus javanensis', 'myophonus glaucinus',
585
+ 'myophonus melanurus', 'myoprocta pratti', 'mystacornis crossleyi',
586
+ 'nandinia binotata', 'cyanomitra cyanolaema', 'oressochen jubatus',
587
+ 'neocossyphus rufus', 'neofelis diardi', 'neofelis nebulosa',
588
+ 'neomorphus geoffroyi', 'neomorphus rufipennis',
589
+ 'delacourella capistrata', 'streptopelia picturata',
590
+ 'nesolagus netscheri', 'nesomys audeberti', 'nesotragus moschatus',
591
+ 'caprimulgus europaeus', 'niltava sumatrana', 'nisaetus nanus',
592
+ 'nothocrax urumutum', 'numida meleagris', 'nyctidromus albicollis',
593
+ 'odontophorus balliviani', 'odontophorus erythrops',
594
+ 'odontophorus gujanensis', 'oenomys hypoxanthus',
595
+ 'ortalis guttata', 'oryx beisa', 'otolemur garnettii',
596
+ 'otus spilocephalus', 'ovis aries', 'oxylabes madagascariensis',
597
+ 'pan troglodytes', 'panthera tigris', 'papio anubis',
598
+ 'papio cynocephalus', 'paraxerus boehmi', 'paraxerus cepapi',
599
+ 'paraxerus lucifer', 'paraxerus ochraceus',
600
+ 'paraxerus vexillarius', 'penelope purpurascens',
601
+ 'penelope superciliaris', 'pernis ptilorhynchus',
602
+ 'petrodromus tetradactylus', 'philander opossum',
603
+ 'philantomba monticola', 'pilherodius pileatus',
604
+ 'xanthomixis apperti', 'pipile cumanensis', 'pipile pipile',
605
+ 'hydrornis guajanus', 'hydrornis schneideri', 'ploceus alienus',
606
+ 'ploceus baglafecht', 'poecilogale albinucha',
607
+ 'pogonocichla stellata', 'polyplectron chalcurum',
608
+ 'erythrogenys mcclellandi', 'potos flavus', 'praomys tullbergi',
609
+ 'presbytis femoralis', 'presbytis thomasi', 'prionodon linsang',
610
+ 'procavia capensis', 'piliocolobus gordonorum',
611
+ 'procyon cancrivorus', 'propithecus candidus',
612
+ 'protoxerus stangeri', 'psophia crepitans', 'psophia leucoptera',
613
+ 'pternistis hildebrandti', 'pternistis leucoscepus',
614
+ 'pteroglossus beauharnaisii', 'ramphastos tucanus',
615
+ 'rattus tiomanicus', 'rhea americana', 'rhizomys sumatrensis',
616
+ 'rhynchocyon cirnei', 'rhynchocyon petersi',
617
+ 'rhynchocyon udzungwensis', 'rollulus rouloul',
618
+ 'rupornis magnirostris', 'ruwenzorornis johnstoni',
619
+ 'saimiri boliviensis', 'salanoia concolor', 'saxicola tectes',
620
+ 'sciurus deppei', 'sciurus granatensis', 'sciurus ignitus',
621
+ 'sciurus spadiceus', 'setifer setosus', 'sheppardia lowei',
622
+ 'spilogale putorius', 'spizaetus ornatus',
623
+ 'stephanoaetus coronatus', 'stigmochelys pardalis',
624
+ 'streptopelia capicola', 'streptopelia lugens',
625
+ 'streptopelia senegalensis', 'amazona oratrix', 'suncus murinus',
626
+ 'sundasciurus hippurus', 'sus barbatus', 'sylvilagus brasiliensis',
627
+ 'tamandua mexicana', 'tapirus bairdii', 'tapirus indicus',
628
+ 'tauraco livingstonii', 'tenrec ecaudatus', 'terpsiphone mutata',
629
+ 'thamnomys venustus', 'thryonomys gregorianus',
630
+ 'thryonomys swinderianus', 'tigrisoma lineatum',
631
+ 'tigrisoma mexicanum', 'tinamus guttatus', 'tinamus tao',
632
+ 'tockus deckeni', 'tockus flavirostris', 'tolypeutes matacus',
633
+ 'tragelaphus imberbis', 'tragulus javanicus', 'tragulus kanchil',
634
+ 'tragulus napu', 'tremarctos ornatus', 'trichys fasciculata',
635
+ 'tupaia glis', 'tupinambis teguixin', 'turdus ignobilis',
636
+ 'turdus olivaceus', 'turdus tephronotus', 'turtur chalcospilos',
637
+ 'turtur tympanistria', 'tyto alba', 'vanellus coronatus',
638
+ 'varanus salvator', 'vicugna pacos', 'viverricula indica',
639
+ 'xenoperdix udzungwensis', 'euxerus erythropus', 'xerus rutilus',
640
+ 'zonotrichia capensis', 'erinaceus europaeus', 'rattus norvegicus'
641
+ ]),
642
+ "subspecies": datasets.ClassLabel(num_classes=8, names=[
643
+ 'sciurus niger cinereus', 'alces alces americanus',
644
+ 'sapajus apella margaritae', 'damaliscus pygargus phillipsi',
645
+ 'alcelaphus buselaphus caama', 'damaliscus lunatus jimela',
646
+ 'equus quagga burchellii', 'zoothera dauma dauma'
647
+ ]),
648
+ "variety": datasets.ClassLabel(num_classes=1, names=[
649
+ 'gallus gallus domesticus'
650
+ ]),
651
+ }
652
+
653
+
654
+ class LILAConfig(datasets.BuilderConfig):
655
+ """Builder Config for LILA"""
656
+ def __init__(self, image_base_url, metadata_url, **kwargs):
657
+ super(LILAConfig, self).__init__(version=datasets.Version("1.0.0"), **kwargs)
658
+ self.image_base_url = image_base_url
659
+ self.metadata_url = metadata_url
660
+
661
+
662
+ class LILA(datasets.GeneratorBasedBuilder):
663
+ """LILA Camera Traps is an aggregate wildlife camera trap dataset for ecological research."""
664
+ VERSION = datasets.Version("1.1.0")
665
+
666
+ BUILDER_CONFIGS = [
667
+ LILAConfig(
668
+ name=row.name,
669
+ image_base_url=row.image_base_url,
670
+ metadata_url=_METADATA_BASE_URL + _LILA_URLS[row.name]
671
+ ) for row in _LILA_SAS_URLS.itertuples()
672
+ ]
673
+
674
+ def _get_features(self) -> datasets.Features:
675
+ if self.config.name == 'Caltech Camera Traps':
676
+ return datasets.Features({
677
+ "file_name": datasets.Value("string"),
678
+ "width": datasets.Value("int32"), "height": datasets.Value("int32"),
679
+ "seq_num_frames": datasets.Value("int32"),
680
+ "date_captured": datasets.Value("string"),
681
+ "seq_id": datasets.Value("string"),
682
+ "location": datasets.Value("string"),
683
+ "rights_holder": datasets.Value("string"),
684
+ "frame_num": datasets.Value("int32"),
685
+ "annotations": datasets.Sequence({
686
+ "taxonomy": _TAXONOMY,
687
+ }),
688
+ "bboxes": datasets.Sequence({
689
+ "taxonomy": _TAXONOMY,
690
+ "bbox": datasets.Sequence(datasets.Value("float32")),
691
+ }),
692
+ "image": datasets.Value("string"),
693
+ })
694
+ elif self.config.name == 'ENA24':
695
+ return datasets.Features({
696
+ "file_name": datasets.Value("string"),
697
+ "width": datasets.Value("int32"), "height": datasets.Value("int32"),
698
+ "annotations": datasets.Sequence({
699
+ "bbox": datasets.Sequence(datasets.Value("float32")),
700
+ "taxonomy": _TAXONOMY,
701
+ }),
702
+ "image": datasets.Value("string"),
703
+ })
704
+ elif self.config.name == 'Missouri Camera Traps':
705
+ return datasets.Features({
706
+ "file_name": datasets.Value("string"),
707
+ "width": datasets.Value("int32"), "height": datasets.Value("int32"),
708
+ "seq_id": datasets.Value("string"), "seq_num_frames": datasets.Value("int32"),
709
+ "frame_num": datasets.Value("int32"),
710
+ "annotations": datasets.Sequence({
711
+ "sequence_level_annotation": datasets.Value("bool"),
712
+ "bbox": datasets.Sequence(datasets.Value("float32")),
713
+ "taxonomy": _TAXONOMY,
714
+ }),
715
+ "image": datasets.Value("string"),
716
+ })
717
+ elif self.config.name == 'NACTI':
718
+ return datasets.Features({
719
+ "file_name": datasets.Value("string"),
720
+ "width": datasets.Value("int32"), "height": datasets.Value("int32"),
721
+ "study": datasets.Value("string"), "location": datasets.Value("string"),
722
+ "annotations": datasets.Sequence({
723
+ "taxonomy": _TAXONOMY,
724
+ }),
725
+ "bboxes": datasets.Sequence({
726
+ "bbox": datasets.Sequence(datasets.Value("float32")),
727
+ }),
728
+ "image": datasets.Value("string"),
729
+ })
730
+ elif self.config.name == 'WCS Camera Traps':
731
+ return datasets.Features({
732
+ "file_name": datasets.Value("string"),
733
+ "width": datasets.Value("int32"), "height": datasets.Value("int32"),
734
+ "wcs_id": datasets.Value("string"), "location": datasets.Value("string"),
735
+ "frame_num": datasets.Value("int32"), "match_level": datasets.Value("int32"),
736
+ "seq_id": datasets.Value("string"), "country_code": datasets.Value("string"),
737
+ "seq_num_frames": datasets.Value("int32"),
738
+ "status": datasets.Value("string"),
739
+ "datetime": datasets.Value("string"),
740
+ "corrupt": datasets.Value("bool"),
741
+ "annotations": datasets.Sequence({
742
+ "count": datasets.Value("int32"),
743
+ "sex": datasets.Value("string"),
744
+ "age": datasets.Value("string"),
745
+ "taxonomy": _TAXONOMY,
746
+ }),
747
+ "bboxes": datasets.Sequence({
748
+ "bbox": datasets.Sequence(datasets.Value("float32")),
749
+ }),
750
+ "image": datasets.Value("string"),
751
+ })
752
+ elif self.config.name == 'Wellington Camera Traps':
753
+ return datasets.Features({
754
+ "file_name": datasets.Value("string"),
755
+ "width": datasets.Value("int32"), "height": datasets.Value("int32"),
756
+ "frame_num": datasets.Value("int32"), "seq_id": datasets.Value("string"),
757
+ "site": datasets.Value("string"), "camera": datasets.Value("string"),
758
+ "datetime": datasets.Value("string"),
759
+ "annotations": datasets.Sequence({
760
+ "taxonomy": _TAXONOMY,
761
+ }),
762
+ "image": datasets.Value("string"),
763
+ })
764
+ elif self.config.name == 'Island Conservation Camera Traps':
765
+ return datasets.Features({
766
+ "file_name": datasets.Value("string"),
767
+ "width": datasets.Value("int32"), "height": datasets.Value("int32"),
768
+ "annotations": datasets.Sequence({
769
+ "bbox": datasets.Sequence(datasets.Value("float32")),
770
+ "taxonomy": _TAXONOMY,
771
+ }),
772
+ "image": datasets.Value("string"),
773
+ })
774
+ elif self.config.name == 'Channel Islands Camera Traps':
775
+ return datasets.Features({
776
+ "file_name": datasets.Value("string"),
777
+ "width": datasets.Value("int32"), "height": datasets.Value("int32"),
778
+ "frame_num": datasets.Value("int32"), "seq_id": datasets.Value("string"),
779
+ "seq_num_frames": datasets.Value("int32"),
780
+ "original_relative_path": datasets.Value("string"),
781
+ "location": datasets.Value("string"),
782
+ "temperature": datasets.Value("string"),
783
+ "annotations": datasets.Sequence({
784
+ "sequence_level_annotation": datasets.Value("bool"),
785
+ "bbox": datasets.Sequence(datasets.Value("float32")),
786
+ "taxonomy": _TAXONOMY,
787
+ }),
788
+ "image": datasets.Value("string"),
789
+ })
790
+ elif self.config.name == 'Idaho Camera Traps':
791
+ return datasets.Features({
792
+ "file_name": datasets.Value("string"),
793
+ "frame_num": datasets.Value("int32"), "seq_id": datasets.Value("string"),
794
+ "seq_num_frames": datasets.Value("int32"),
795
+ "original_relative_path": datasets.Value("string"),
796
+ "datetime": datasets.Value("string"),
797
+ "location": datasets.Value("string"),
798
+ "annotations": datasets.Sequence({
799
+ "sequence_level_annotation": datasets.Value("bool"),
800
+ "taxonomy": _TAXONOMY,
801
+ }),
802
+ "image": datasets.Value("string"),
803
+ })
804
+ elif self.config.name == 'Snapshot Serengeti':
805
+ return datasets.Features({
806
+ "file_name": datasets.Value("string"),
807
+ "frame_num": datasets.Value("int32"), "seq_id": datasets.Value("string"),
808
+ "width": datasets.Value("int32"), "height": datasets.Value("int32"),
809
+ "seq_num_frames": datasets.Value("int32"),
810
+ "datetime": datasets.Value("string"),
811
+ "corrupt": datasets.Value("bool"),
812
+ "location": datasets.Value("string"),
813
+ "annotations": datasets.Sequence({
814
+ "sequence_level_annotation": datasets.Value("bool"),
815
+ "seq_id": datasets.Value("string"),
816
+ "season": datasets.Value("string"),
817
+ "datetime": datasets.Value("string"),
818
+ "subject_id": datasets.Value("string"),
819
+ "count": datasets.Value("string"),
820
+ "standing": datasets.Value("float32"),
821
+ "resting": datasets.Value("float32"),
822
+ "moving": datasets.Value("float32"),
823
+ "interacting": datasets.Value("float32"),
824
+ "young_present": datasets.Value("float32"),
825
+ "location": datasets.Value("string"),
826
+ "taxonomy": _TAXONOMY,
827
+ }),
828
+ "bboxes": datasets.Sequence({
829
+ "bbox": datasets.Sequence(datasets.Value("float32")),
830
+ }),
831
+ "image": datasets.Value("string"),
832
+ })
833
+ elif self.config.name in [
834
+ 'Snapshot Karoo', 'Snapshot Kgalagadi', 'Snapshot Enonkishu', 'Snapshot Camdeboo',
835
+ 'Snapshot Mountain Zebra', 'Snapshot Kruger'
836
+ ]:
837
+ return datasets.Features({
838
+ "file_name": datasets.Value("string"),
839
+ "frame_num": datasets.Value("int32"), "seq_id": datasets.Value("string"),
840
+ "width": datasets.Value("int32"), "height": datasets.Value("int32"),
841
+ "seq_num_frames": datasets.Value("int32"),
842
+ "datetime": datasets.Value("string"),
843
+ "corrupt": datasets.Value("bool"),
844
+ "location": datasets.Value("string"),
845
+ "annotations": datasets.Sequence({
846
+ "sequence_level_annotation": datasets.Value("bool"),
847
+ "seq_id": datasets.Value("string"),
848
+ "season": datasets.Value("string"),
849
+ "datetime": datasets.Value("string"),
850
+ "subject_id": datasets.Value("string"),
851
+ "count": datasets.Value("string"),
852
+ "standing": datasets.Value("float32"),
853
+ "resting": datasets.Value("float32"),
854
+ "moving": datasets.Value("float32"),
855
+ "interacting": datasets.Value("float32"),
856
+ "young_present": datasets.Value("float32"),
857
+ "location": datasets.Value("string"),
858
+ "taxonomy": _TAXONOMY,
859
+ }),
860
+ "image": datasets.Value("string"),
861
+ })
862
+ elif self.config.name == 'SWG Camera Traps':
863
+ return datasets.Features({
864
+ "file_name": datasets.Value("string"),
865
+ "width": datasets.Value("int32"), "height": datasets.Value("int32"),
866
+ "location": datasets.Value("string"),
867
+ "frame_num": datasets.Value("int32"),
868
+ "seq_id": datasets.Value("string"),
869
+ "seq_num_frames": datasets.Value("int32"),
870
+ "datetime": datasets.Value("string"),
871
+ "corrupt": datasets.Value("bool"),
872
+ "annotations": datasets.Sequence({
873
+ "sequence_level_annotation": datasets.Value("bool"),
874
+ "taxonomy": _TAXONOMY,
875
+ }),
876
+ "bboxes": datasets.Sequence({
877
+ "sequence_level_annotation": datasets.Value("bool"),
878
+ "bbox": datasets.Sequence(datasets.Value("float32")),
879
+ "taxonomy": _TAXONOMY,
880
+ }),
881
+ "image": datasets.Value("string"),
882
+ })
883
+ elif self.config.name == 'Orinoquia Camera Traps':
884
+ return datasets.Features({
885
+ "file_name": datasets.Value("string"),
886
+ "frame_num": datasets.Value("int32"), "seq_id": datasets.Value("string"),
887
+ "seq_num_frames": datasets.Value("int32"), "datetime": datasets.Value("string"),
888
+ "location": datasets.Value("string"),
889
+ "annotations": datasets.Sequence({
890
+ "sequence_level_annotation": datasets.Value("bool"),
891
+ "taxonomy": _TAXONOMY,
892
+ }),
893
+ "image": datasets.Value("string"),
894
+ })
895
+
896
+ def _info(self):
897
+ features = self._get_features()
898
+
899
+ return datasets.DatasetInfo(
900
+ description=_DESCRIPTION,
901
+ features=features,
902
+ homepage=_HOMEPAGE,
903
+ citation=_LILA_CITATIONS[self.config.name],
904
+ )
905
+
906
+ def _split_generators(self, dl_manager):
907
+ archive_path = dl_manager.download_and_extract(self.config.metadata_url)
908
+ if archive_path.endswith(".zip") or os.path.isdir(archive_path):
909
+ archive_path = os.path.join(archive_path, os.listdir(archive_path)[0])
910
+
911
+ return [
912
+ datasets.SplitGenerator(
913
+ name=datasets.Split.TRAIN,
914
+ gen_kwargs={
915
+ "filepath": archive_path,
916
+ "split": "train",
917
+ },
918
+ ),
919
+ ]
920
+
921
+ def _generate_examples(self, filepath, split):
922
+ with open(filepath) as f:
923
+ for line in f:
924
+ example = json.loads(line)
925
+ image_url = f"{self.config.image_base_url}/{example['file_name']}"
926
+ yield example["file_name"], {
927
+ **example,
928
+ "image": image_url
929
+ }