Exposing Image Classifier Shortcuts with Counterfactual Frequency (CoF) Tables

TL;DR

This paper introduces Counterfactual Frequency (CoF) tables, a novel method for aggregating explanations to identify shortcuts in image classifiers, improving interpretability and detection of spurious patterns.

Contribution

The paper proposes CoF tables, a new approach that aggregates instance explanations into global insights to expose shortcuts in image classification models.

Findings

CoF tables effectively reveal shortcuts learned by models.

Application across multiple datasets demonstrates utility.

Facilitates easier detection of spurious patterns.

Abstract

The rise of deep learning in image classification has brought unprecedented accuracy but also highlighted a key issue: the use of 'shortcuts' by models. Such shortcuts are easy-to-learn patterns from the training data that fail to generalise to new data. Examples include the use of a copyright watermark to recognise horses, snowy background to recognise huskies, or ink markings to detect malignant skin lesions. The explainable AI (XAI) community has suggested using instance-level explanations to detect shortcuts without external data, but this requires the examination of many explanations to confirm the presence of such shortcuts, making it a labour-intensive process. To address these challenges, we introduce Counterfactual Frequency (CoF) tables, a novel approach that aggregates instance-based explanations into global insights, and exposes shortcuts. The aggregation implies the need for some semantic concepts to be used in the explanations, which we solve by labelling the segments of an image. We demonstrate the utility of CoF tables across several datasets, revealing the shortcuts learned from them.