TDLS: A Top-Down Layer Searching Algorithm for Generating Counterfactual Visual Explanation

TL;DR

This paper introduces TDLS, a top-down layer searching algorithm that generates flexible and efficient counterfactual visual explanations for fine-grained image classification, enhancing transparency and interpretability of AI models.

Contribution

The paper proposes a novel top-down layer searching algorithm (TDLS) for generating counterfactual visual explanations in image classification tasks, demonstrating improved flexibility and efficiency.

Findings

TDLS provides more flexible counterfactual explanations.

The method is efficient on VGG-16 model.

Effective for fine-grained image classification.

Abstract

Explanation of AI, as well as fairness of algorithms' decisions and the transparency of the decision model, are becoming more and more important. And it is crucial to design effective and human-friendly techniques when opening the black-box model. Counterfactual conforms to the human way of thinking and provides a human-friendly explanation, and its corresponding explanation algorithm refers to a strategic alternation of a given data point so that its model output is "counter-facted", i.e. the prediction is reverted. In this paper, we adapt counterfactual explanation over fine-grained image classification problem. We demonstrated an adaptive method that could give a counterfactual explanation by showing the composed counterfactual feature map using top-down layer searching algorithm (TDLS). We have proved that our TDLS algorithm could provide more flexible counterfactual visual explanation in an efficient way using VGG-16 model on Caltech-UCSD Birds 200 dataset. At the end, we discussed several applicable scenarios of counterfactual visual explanations.