We develop a novel post-hoc visual explanation method called Score-CAM, which is the first gradient-free CAM-based visualization method that achieves better visual performance (state-of-the-art).
Paper: Score-CAM: Score-Weighted Visual Explanations for Convolutional Neural Networks, appeared at IEEE CVPR 2020 Workshop on Fair, Data Efficient and Trusted Computer Vision. Our paper has been cited nearly 100!
Demo: You can run an example via Colab
2021.4.03: Merged into jacobgil/pytorch-grad-cam (2.8K Stars).
2020.8.18: Merged into PaddlePaddle/InterpretDL, a toolkit for PaddlePaddle models.
2020.7.11: Merged into keisen/tf-keras-vis, written in Tensorflow.
2020.5.11: Merged into utkuozbulak/pytorch-cnn-visualizations (5.7K Stars).
2020.3.24: Merged into frgfm/torch-cam, a wonderful library that supports multiple CAM-based methods.
If you find this work or code is helpful in your research, please cite our work:
@inproceedings{wang2020score,
title={Score-CAM: Score-weighted visual explanations for convolutional neural networks},
author={Wang, Haofan and Wang, Zifan and Du, Mengnan and Yang, Fan and Zhang, Zijian and Ding, Sirui and Mardziel, Piotr and Hu, Xia},
booktitle={Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops},
pages={24--25},
year={2020}
}
Utils are built on flashtorch, thanks for releasing this great work!
If you have any questions, feel free to open an issue or directly contact me via: haofanw@andrew.cmu.edu