The official code for V2L: Leveraging Vision and Vision-language Models into Large-scale Product Retrieval (CVPR 2022 Challenge Rank 1 Solution)
FGVC9: eBay eProduct Visual Search Challenge 2022
If you are interested in the code, please write a request email (with your purpose and your personal information) to [email protected].
The codebase is based on our former work 1st Place Solution of the Facebook AI Image Similarity Challenge and 3rd Place Solution of the Facebook AI Image Similarity Challenge.
@misc{wang2022v2l,
title={V$^2$L: Leveraging Vision and Vision-language Models into Large-scale Product Retrieval},
author={Wenhao Wang and Yifan Sun and Zongxin Yang and Yi Yang},
year={2022},
eprint={2207.12994},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
@article{wang2021bag,
title={Bag of tricks and a strong baseline for image copy detection},
author={Wang, Wenhao and Zhang, Weipu and Sun, Yifan and Yang, Yi},
journal={arXiv preprint arXiv:2111.08004},
year={2021}
}
@article{wang2021d,
title={D\^{} 2LV: A Data-Driven and Local-Verification Approach for Image Copy Detection},
author={Wang, Wenhao and Sun, Yifan and Zhang, Weipu and Yang, Yi},
journal={arXiv preprint arXiv:2111.07090},
year={2021}
}