This repository contains the core code to implement the idea of our paper:
Discover and Mitigate Multiple Biased Subgroups in Image Classifiers (CVPR 2024)
- python >= 3.6.5
- pytorch == 1.7.x
- numpy >= 1.15.4
- clip-retrieval >= 2.0
If you find the idea or code useful for your research, please consider citing our paper:
@inproceedings{zhang2024DIM,
author={Zeliang Zhang and Mingqian Feng and Zhiheng Li and Chenliang Xu},
booktitle = {The Thirty-Fourth IEEE/CVF Conference on Computer Vision and Pattern Recognition},
title = {Discover and Mitigate Multiple Biased Subgroups in Image Classifiers},
year = {2024},
}
Questions and suggestions can be sent to hust0426@gmail.com.