CACR

CACR checkpoint repository

For implementation details, please refer to our Github Repo

ImageNet pretrained, performance of linear classification on ImageNet

<table><tbody> <!-- START TABLE --> <!-- TABLE HEADER --> <th align="left">model</th> <th align="left">pretrain<br/>epochs</th> <th align="left">linear<br/>acc</th> <!-- TABLE BODY --> <tr> <td align="left">ResNet50</td> <td align="left">1000</td> <td align="left">74.7</td> </tr> <tr> <td align="left">ViT-Base</td> <td align="left">300</td> <td align="left">77.1</td> </tr> </tbody></table>

ImageNet pretrained, performance of linear classification on 20 Image in the Wild datasets

Please feel free to check our learned representation performance in Image in the Wild Challenge.

Citation

Please cite our work if you find it is helpful. Thank you!

@article{
  zheng2023contrastive,
  title={Contrastive Attraction and Contrastive Repulsion for Representation Learning},
  author={Huangjie Zheng and Xu Chen and Jiangchao Yao and Hongxia Yang and Chunyuan Li and Ya Zhang and Hao Zhang and Ivor Tsang and Jingren Zhou and Mingyuan Zhou},
  journal={Transactions on Machine Learning Research},
  issn={2835-8856},
  year={2023},
  url={https://openreview.net/forum?id=f39UIDkwwc},
}