XRAI: Better Attributions Through Regions
This work addresses the need for more interpretable AI models, but it is incremental as it builds upon existing integrated gradients.
The authors tackled the problem of improving saliency methods for understanding deep neural networks by introducing XRAI, a region-based attribution method, and showed it produces better results than other methods on common models and the ImageNet dataset.
Saliency methods can aid understanding of deep neural networks. Recent years have witnessed many improvements to saliency methods, as well as new ways for evaluating them. In this paper, we 1) present a novel region-based attribution method, XRAI, that builds upon integrated gradients (Sundararajan et al. 2017), 2) introduce evaluation methods for empirically assessing the quality of image-based saliency maps (Performance Information Curves (PICs)), and 3) contribute an axiom-based sanity check for attribution methods. Through empirical experiments and example results, we show that XRAI produces better results than other saliency methods for common models and the ImageNet dataset.