Awesome

Co-advise

This is the official implementation of CVPR2022 paper "Co-advise: Cross Inductive Bias Distillation"

Prepare

Install RedNet following https://github.com/d-li14/involution

Install the rest packages following https://github.com/microsoft/Swin-Transformer

Training

bash train.sh

Noted that mix precision should be closed, because involution is not supported.

Pretrained Model

The checkpoints can be found at Goolge Drive

Citation

@InProceedings{Ren_coadvise_2022_CVPR,
    author    = {Sucheng Ren, Zhengqi Gao, Tianyu Hua, Zihui Xue, Yonglong Tian, Shengfeng He, Hang Zhao},
    title     = {Co-advise: Cross Inductive Bias Distillation},
    booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
    month     = {June},
    year      = {2022},
}

If you have any questions, feel free to email Sucheng Ren :) (oliverrensu@gmail.com)