LeqiShen/DiscoVLA

This dataset (MSRVTT_PseudoImageCaptions.json) is released as part of our CVPR 2025 paper:

DiscoVLA: Discrepancy Reduction in Vision, Language, and Alignment for Parameter-Efficient Video-Text Retrieval

ArXiv: https://arxiv.org/abs/2506.08887

Github: https://github.com/LunarShen/DsicoVLA

Citation

@inproceedings{shen2025discovla,
  title={DiscoVLA: Discrepancy Reduction in Vision, Language, and Alignment for Parameter-Efficient Video-Text Retrieval},
  author={Shen, Leqi and Gong, Guoqiang and Hao, Tianxiang and He, Tao and Zhang, Yifeng and Liu, Pengzhang and Zhao, Sicheng and Han, Jungong and Ding, Guiguang},
  booktitle={Proceedings of the Computer Vision and Pattern Recognition Conference},
  pages={19702--19712},
  year={2025}
}