How to Evaluate and Refine your CAM

TL;DR

This paper introduces a synthetic dataset for evaluating CAMs, proposes a new composite metric ARCC for better faithfulness assessment, and presents RefineCAM for higher-resolution attribution maps, improving interpretability.

Contribution

It provides a ground-truth dataset for CAM evaluation, a new metric ARCC for more reliable explanation assessment, and a method RefineCAM for high-resolution attribution maps.

Findings

ARCC more reliably identifies faithful explanations.

RefineCAM produces higher-resolution attribution maps.

RefineCAM outperforms existing CAM methods in evaluations.

Abstract

Class attribution maps (CAMs) provide local explanations for the decisions of convolutional neural networks. While widely used in practice, the evaluation of CAMs remains challenging due to the lack of ground-truth explanations, making it difficult to evaluate the soundness of existing metrics. Independently, most commonly used CAM methods produce low-resolution attribution maps, which limits their usefulness for detailed interpretability. To address the evaluation challenge, we introduce a synthetic dataset with ground-truth attributions that enables a rigorous comparison of CAM evaluation metrics. Using this dataset, we analyze existing metrics and propose ARCC, a new composite metric that more reliably identifies faithful explanations. To address the low resolution issue, we introduce RefineCAM, a method that produces high-resolution attribution maps by aggregating CAMs across multiple network layers. Our results show that RefineCAM consistently outperforms existing methods according to the proposed evaluation.