VALE: A Multimodal Visual and Language Explanation Framework for Image Classifiers using eXplainable AI and Language Models

TL;DR

This work tackles the interpretability of image classifiers by introducing VALE, a multimodal XAI framework that fuses SHAP-based visual explanations, Segment Anything Model segmentation, and vision-language model-generated textual explanations. The approach bridges the semantic gap between machine outputs and human understanding, validated on ImageNet and underwater SONAR datasets with quantitative and BLEU-based qualitative assessments. Key contributions include a novel multimodal explainer, seamless integration of segmentation and captioning models, and demonstration of prompt engineering to tailor explanations for real-world tasks. The results indicate enhanced interpretability and potential for deploying image classifiers in high-risk domains, with future work focusing on expanding XAI tool integration and broader domain applicability.

Abstract

Deep Neural Networks (DNNs) have revolutionized various fields by enabling task automation and reducing human error. However, their internal workings and decision-making processes remain obscure due to their black box nature. Consequently, the lack of interpretability limits the application of these models in high-risk scenarios. To address this issue, the emerging field of eXplainable Artificial Intelligence (XAI) aims to explain and interpret the inner workings of DNNs. Despite advancements, XAI faces challenges such as the semantic gap between machine and human understanding, the trade-off between interpretability and performance, and the need for context-specific explanations. To overcome these limitations, we propose a novel multimodal framework named VALE Visual and Language Explanation. VALE integrates explainable AI techniques with advanced language models to provide comprehensive explanations. This framework utilizes visual explanations from XAI tools, an advanced zero-shot image segmentation model, and a visual language model to generate corresponding textual explanations. By combining visual and textual explanations, VALE bridges the semantic gap between machine outputs and human interpretation, delivering results that are more comprehensible to users. In this paper, we conduct a pilot study of the VALE framework for image classification tasks. Specifically, Shapley Additive Explanations (SHAP) are used to identify the most influential regions in classified images. The object of interest is then extracted using the Segment Anything Model (SAM), and explanations are generated using state-of-the-art pre-trained Vision-Language Models (VLMs). Extensive experimental studies are performed on two datasets: the ImageNet dataset and a custom underwater SONAR image dataset, demonstrating VALEs real-world applicability in underwater image classification.

Figures (3)

• Figure 1: Architecture of VALE: Visual and Language Explainer framework.
• Figure 2: Output from the SHAP explainer (Explanation), the coordinate with the highest SHAP value (ROI - represented with a magenta star), and the generated mask for Bald Eagle.
• Figure 3: Explanation for Bald Eagle image illustrating the pipeline.