Benchmarking XAI Explanations with Human-Aligned Evaluations
This work addresses the need for reliable and consistent human-aligned evaluations of XAI methods, which is incremental as it builds on existing XAI research by providing a new benchmark and scoring system.
The paper tackles the problem of evaluating explainable AI (XAI) techniques in computer vision by introducing PASTA, a human-centric framework that includes a large-scale benchmark dataset and an automated scoring method, resulting in scalable and human-aligned evaluations.
We introduce PASTA (Perceptual Assessment System for explanaTion of Artificial Intelligence), a novel human-centric framework for evaluating eXplainable AI (XAI) techniques in computer vision. Our first contribution is the creation of the PASTA-dataset, the first large-scale benchmark that spans a diverse set of models and both saliency-based and concept-based explanation methods. This dataset enables robust, comparative analysis of XAI techniques based on human judgment. Our second contribution is an automated, data-driven benchmark that predicts human preferences using the PASTA-dataset. This scoring called PASTA-score method offers scalable, reliable, and consistent evaluation aligned with human perception. Additionally, our benchmark allows for comparisons between explanations across different modalities, an aspect previously unaddressed. We then propose to apply our scoring method to probe the interpretability of existing models and to build more human interpretable XAI methods.