To appear at IEEE CVPR 2020 Workshop on Fair, Data Efficient and Trusted Computer Vision.
In this paper, we develop a novel post-hoc visual explanation method called Score-CAM based on class activation mapping. Score-CAM is a gradient-free visualization method, extended from Grad-CAM and Grad-CAM++. It achieves better visual performance and fairness for interpreting the decision making process.
Paper: Score-CAM: Score-Weighted Visual Explanations for Convolutional Neural Networks (Haofan Wang, Zifan Wang, Mengnan Du, Fan Yang, Zijian Zhang, Sirui Ding, Piotr Mardziel and Xia Hu.)
2020.8.18: Score-CAM has been merged into PaddlePaddle/InterpretDL.
2020.7.11: Score-CAM has been merged into keisen/tf-keras-vis.
2020.5.11: Score-CAM has been merged into utkuozbulak/pytorch-cnn-visualizations.
2020.3.24: Score-CAM has been merged into frgfm/torch-cam.
If you find this work or code is helpful in your research, please cite and star:
@InProceedings{Wang_2020_CVPR_Workshops,
author = {Wang, Haofan and Wang, Zifan and Du, Mengnan and Yang, Fan and Zhang, Zijian and Ding, Sirui and Mardziel, Piotr and Hu, Xia},
title = {Score-CAM: Score-Weighted Visual Explanations for Convolutional Neural Networks},
booktitle = {The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
month = {June},
year = {2020}
}
Utils are built on flashtorch, thanks for releasing this great work!
If you have any questions, feel free to contact me via: haofanw@andrew.cmu.edu