Jerry00917 / samshap

Code for the paper "Explain Any Concept: Segment Anything Meets Concept-Based Explanation". Poster @ NeurIPS 2023

Explain Any Concept: Segment Anything Meets Concept-Based Explanation (EAC) Poster @ NeurIPS 2023

Code for the paper "Explain Any Concept: Segment Anything Meets Concept-Based Explanation".

Citation

Please cite the paper as follows if you use the data or code from Samshap:

@inproceedings{
sun2023explain,
title={Explain Any Concept: Segment Anything Meets Concept-Based Explanation},
author={Ao Sun and Pingchuan Ma and Yuanyuan Yuan and Shuai Wang},
booktitle={Thirty-seventh Conference on Neural Information Processing Systems},
year={2023},
url={https://openreview.net/forum?id=X6TBBsz9qi}
}

Contact

Please reach out to us if you have any questions or suggestions. You can send an email to [e-mail].

Overview

Here is an overview of our work, and you can find more in our Paper.

Our EAC approach generates high accurate and human-understandable post-hoc explanations.

Downloading the SAM backbone

We use ViT-H as our default SAM model. For downloading the pre-train model and installation dependencies, please refer SAM repo.

Explain a hummingbird on your local pre-trained ResNet-50!

Simply run the following command:

python demo_samshap.py