The VILD Dataset (VIdeo and Long-Description)
This dataset is proposed from VideoCLIP-XL. We establish an automatic data collection system, designed to aggregate sufficient and high-quality pairs from multiple data sources. We have successfully collected over 2M (VIdeo, Long Description) pairs, denoted as our VILD dataset.
Format
{
"short_captions": [
"...",
],
"long_captions": [
"...",
],
"video_id": "..."
}
{
.....
},
.....
Source
@misc{wang2024videoclipxladvancinglongdescription,
title={VideoCLIP-XL: Advancing Long Description Understanding for Video CLIP Models},
author={Jiapeng Wang and Chengyu Wang and Kunzhe Huang and Jun Huang and Lianwen Jin},
year={2024},
eprint={2410.00741},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2410.00741},
}