Explaning with trees: interpreting CNNs using hierarchies

TL;DR

This paper introduces xAiTrees, a hierarchical segmentation framework that provides faithful, multiscale explanations of CNNs, improving interpretability and aiding bias detection in neural network reasoning.

Contribution

The paper presents a novel hierarchical segmentation approach for CNN interpretability that maintains model fidelity and offers multiscale explanations, surpassing existing xAI methods.

Findings

xAiTrees delivers highly interpretable explanations.

The method outperforms traditional xAI techniques.

It helps identify biases in neural network reasoning.

Abstract

Challenges persist in providing interpretable explanations for neural network reasoning in explainable AI (xAI). Existing methods like Integrated Gradients produce noisy maps, and LIME, while intuitive, may deviate from the model's reasoning. We introduce a framework that uses hierarchical segmentation techniques for faithful and interpretable explanations of Convolutional Neural Networks (CNNs). Our method constructs model-based hierarchical segmentations that maintain the model's reasoning fidelity and allows both human-centric and model-centric segmentation. This approach offers multiscale explanations, aiding bias identification and enhancing understanding of neural network decision-making. Experiments show that our framework, xAiTrees, delivers highly interpretable and faithful model explanations, not only surpassing traditional xAI methods but shedding new light on a novel approach to enhancing xAI interpretability.