Visual quality inspection systems, crucial in sectors like manufacturing and logistics, employ computer vision and machine learning for precise, rapid defect detection. However, their unexplained nature can hinder trust, error identification, and system improvement. This paper presents a framework to bolster visual quality inspection by using CAM-based explanations to refine semantic segmentation models. Our approach consists of 1) Model Training, 2) XAI-based Model Explanation, 3) XAI Evaluation, and 4) Annotation Augmentation for Model Enhancement, informed by explanations and expert insights. Evaluations show XAI-enhanced models surpass original DeepLabv3-ResNet101 models, especially in intricate object segmentation.

翻译：视觉质量检验系统在制造业和物流等关键领域采用计算机视觉与机器学习技术，实现精准快速的缺陷检测。然而其不可解释性可能阻碍信任建立、错误识别与系统优化。本文提出一种框架，通过基于CAM的解释方法优化语义分割模型，以增强视觉质量检验能力。该方法包含四个阶段：1）模型训练，2）基于XAI的模型解释，3）XAI评估，以及4）基于解释与专家见解的标注增强模型优化。评估结果表明，经XAI增强的模型在复杂目标分割任务中显著优于原始DeepLabv3-ResNet101模型。