Beyond Pixels: Enhancing LIME with Hierarchical Features and Segmentation Foundation Models

TL;DR

This paper introduces DSEG-LIME, an enhanced version of LIME that uses data-driven segmentation and hierarchical features from foundation models to produce more accurate and human-aligned explanations for vision models.

Contribution

The paper proposes DSEG-LIME, integrating foundation model-based segmentation and hierarchical granularity control to improve explanation quality in vision models.

Findings

DSEG outperforms existing LIME variants on XAI metrics

Enhanced explanations better align with human-recognized concepts

Improved interpretability for pre-trained ImageNet models

Abstract

LIME (Local Interpretable Model-agnostic Explanations) is a popular XAI framework for unraveling decision-making processes in vision machine-learning models. The technique utilizes image segmentation methods to identify fixed regions for calculating feature importance scores as explanations. Therefore, poor segmentation can weaken the explanation and reduce the importance of segments, ultimately affecting the overall clarity of interpretation. To address these challenges, we introduce the DSEG-LIME (Data-Driven Segmentation LIME) framework, featuring: i) a data-driven segmentation for human-recognized feature generation by foundation model integration, and ii) a user-steered granularity in the hierarchical segmentation procedure through composition. Our findings demonstrate that DSEG outperforms on several XAI metrics on pre-trained ImageNet models and improves the alignment of explanations with human-recognized concepts. The code is available under: https://github. com/patrick-knab/DSEG-LIME