Axiomatic Explanations for Visual Search, Retrieval, and Similarity Learning

TL;DR

This paper introduces a formal axiomatic framework based on fair credit assignment for explaining visual search and similarity models, extending existing methods and improving explanation consistency and efficiency.

Contribution

It provides a unique axiomatic solution that generalizes and extends multiple explainability techniques for visual search and similarity models, addressing fairness and computational efficiency.

Findings

The formalism unifies existing explanation methods under a fair credit assignment framework.

Extensions of CAM, GradCAM, LIME, SHAP, and SBSM to search engines are proposed.

A fast kernel-based method for estimating Shapley-Taylor indices is introduced.

Abstract

Visual search, recommendation, and contrastive similarity learning power technologies that impact billions of users worldwide. Modern model architectures can be complex and difficult to interpret, and there are several competing techniques one can use to explain a search engine's behavior. We show that the theory of fair credit assignment provides a $\textit{unique}$ axiomatic solution that generalizes several existing recommendation- and metric-explainability techniques in the literature. Using this formalism, we show when existing approaches violate "fairness" and derive methods that sidestep these shortcomings and naturally handle counterfactual information. More specifically, we show existing approaches implicitly approximate second-order Shapley-Taylor indices and extend CAM, GradCAM, LIME, SHAP, SBSM, and other methods to search engines. These extensions can extract pairwise correspondences between images from trained $\textit{opaque-box}$ models. We also introduce a fast kernel-based method for estimating Shapley-Taylor indices that require orders of magnitude fewer function evaluations to converge. Finally, we show that these game-theoretic measures yield more consistent explanations for image similarity architectures.