TbExplain: A Text-Based Explanation Method for Scene Classification Models with the Statistical Prediction Correction

TY - GEN

T1 - TbExplain

T2 - 1st Workshop on Governance, Understanding and Integration of Data for Effective and Responsible AI, GUIDE-AI 2024

AU - Aminimehr, Amirhossein

AU - Khani, Pouya

AU - Molaei, Amirali

AU - Kazemeini, Amirmohammad

AU - Cambria, Erik

PY - 2024/6/9

Y1 - 2024/6/9

N2 - Heatmaps are common tools in Explainable Artificial Intelligence (XAI) field, but they are not without imperfections; E.g., non-expert users may not grasp the underlying rationale of heatmaps, wherein pixels relevant to the model's prediction are highlighted through distinct intensities or colors. Moreover, objects and regions of the input image that are relevant to the model prediction are frequently not entirely differentiated by heatmaps. In this paper, we propose a framework called TbExplain that employs XAI techniques and a pre-trained object detector to present text-based explanations of scene classification models. Moreover, TbExplain incorporates a novel method to correct predictions and textually explain them based on the statistics of objects in the input image when the initial prediction is unreliable. To assess the trustworthiness and validity of the text-based explanations, we conducted a qualitative experiment, and the findings indicated that these explanations are sufficiently reliable. Furthermore, our quantitative and qualitative experiments on TbExplain with scene classification datasets reveal an improvement in classification accuracy over ResNet variants.

AB - Heatmaps are common tools in Explainable Artificial Intelligence (XAI) field, but they are not without imperfections; E.g., non-expert users may not grasp the underlying rationale of heatmaps, wherein pixels relevant to the model's prediction are highlighted through distinct intensities or colors. Moreover, objects and regions of the input image that are relevant to the model prediction are frequently not entirely differentiated by heatmaps. In this paper, we propose a framework called TbExplain that employs XAI techniques and a pre-trained object detector to present text-based explanations of scene classification models. Moreover, TbExplain incorporates a novel method to correct predictions and textually explain them based on the statistics of objects in the input image when the initial prediction is unreliable. To assess the trustworthiness and validity of the text-based explanations, we conducted a qualitative experiment, and the findings indicated that these explanations are sufficiently reliable. Furthermore, our quantitative and qualitative experiments on TbExplain with scene classification datasets reveal an improvement in classification accuracy over ResNet variants.

KW - explainable artificial intelligence

KW - image classification

KW - interpretability

KW - scene recognition

UR - https://www.scopus.com/pages/publications/85197233364

U2 - 10.1145/3665601.3669841

DO - 10.1145/3665601.3669841

M3 - Article in proceedings

AN - SCOPUS:85197233364

SP - 54

EP - 60

BT - GUIDE-AI '24

PB - Association for Computing Machinery

Y2 - 14 June 2024

ER -