Summary of Less Is More: Fewer Interpretable Region Via Submodular Subset Selection, by Ruoyu Chen et al.
Less is More: Fewer Interpretable Region via Submodular Subset Selection
by Ruoyu Chen, Hua Zhang, Siyuan Liang, Jingzhi Li, Xiaochun Cao
First submitted to arxiv on: 14 Feb 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed method re-models image attribution as a submodular subset selection problem to enhance model interpretability using fewer regions. A novel submodular function is constructed to discover more accurate small interpretation regions, and four constraints are imposed on the selection of sub-regions to assess their importance. Theoretical analysis confirms that the proposed function is indeed submodular. Experimental results show that the method outperforms state-of-the-art methods on three datasets, achieving gains in deletion and insertion scores for correctly predicted samples, as well as improvements in average highest confidence and insertion score for incorrectly predicted samples. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps us understand how machines make decisions by identifying important regions in images. Right now, existing methods have some problems: they can be wrong about small parts of the image, and they’re not good at explaining why they made a mistake. To fix these issues, this study turns the problem into a new kind of optimization problem that focuses on finding smaller areas in the image that are important. The team also adds four rules to make sure the method is reliable. They tested their approach on three different types of images and found it works better than other methods. |
Keywords
* Artificial intelligence * Optimization
