Explaining Image Classifiers by Counterfactual Generation

TL;DR

This paper introduces a method for explaining image classifiers by generating counterfactuals through generative models, identifying image regions that most influence the classifier's decisions with more accuracy and fewer artifacts.

Contribution

It proposes a novel counterfactual generation approach for saliency maps that outperforms previous methods by producing more relevant and artifact-free explanations.

Findings

Produces more compact and relevant saliency maps

Generates plausible in-fills by conditioning on the rest of the image

Outperforms ad-hoc in-filling approaches in explanation quality

Abstract

When an image classifier makes a prediction, which parts of the image are relevant and why? We can rephrase this question to ask: which parts of the image, if they were not seen by the classifier, would most change its decision? Producing an answer requires marginalizing over images that could have been seen but weren't. We can sample plausible image in-fills by conditioning a generative model on the rest of the image. We then optimize to find the image regions that most change the classifier's decision after in-fill. Our approach contrasts with ad-hoc in-filling approaches, such as blurring or injecting noise, which generate inputs far from the data distribution, and ignore informative relationships between different parts of the image. Our method produces more compact and relevant saliency maps, with fewer artifacts compared to previous methods.