LR-IAD：基于逻辑推理的无掩码工业异常检测
LR-IAD:Mask-Free Industrial Anomaly Detection with Logical Reasoning
AI总结：
针对工业异常检测中依赖掩码标注和类别不平衡问题，提出无掩码推理框架LR-IAD，结合动态奖励函数与CoT、GRPO机制直接从原始图像检测缺陷，在MVTec-AD和VisA数据集上准确率分别提升36%和16%。
摘要：
工业异常检测（IAD）通过识别缺陷来确保产品质量，至关重要。传统方法如特征嵌入和基于重建的方法需要大型数据集，且在可扩展性方面存在困难。现有的视觉语言模型（VLMs）和多模态大语言模型（MLLMs）解决了一些局限性，但依赖掩码标注，导致实施成本高并产生假阳性。此外，像 MVTec-AD 和 VisA 这样的工业数据集存在严重的类别不平衡问题，缺陷样本分别仅占总数据的 23.8% 和 11.1%。为了应对这些挑战，我们提出了一种奖励函数，在训练期间动态优先处理罕见的缺陷模式以解决类别不平衡问题。我们还引入了一个无掩码推理框架，使用思维链（CoT）和组相对策略优化（GRPO）机制，能够直接从原始图像进行异常检测，无需标注掩码。该方法为缺陷定位生成可解释的逐步解释。我们的方法实现了最先进的性能，在 MVTec-AD 上准确率比先前方法高出 36%，在 VisA 上高出 16%。通过消除掩码依赖、降低成本并提供可解释的输出，这项工作推进了工业异常检测，并支持制造业中可扩展的质量控制。复现实验的代码可在 https://github.com/LilaKen/LR-IAD 获取。
英文摘要：
Industrial Anomaly Detection (IAD) is critical for ensuring product quality by identifying defects. Traditional methods such as feature embedding and reconstruction-based approaches require large datasets and struggle with scalability. Existing vision-language models (VLMs) and Multimodal Large Language Models (MLLMs) address some limitations but rely on mask annotations, leading to high implementation costs and false positives. Additionally, industrial datasets like MVTec-AD and VisA suffer from severe class imbalance, with defect samples constituting only 23.8% and 11.1% of total data respectively. To address these challenges, we propose a reward function that dynamically prioritizes rare defect patterns during training to handle class imbalance. We also introduce a mask-free reasoning framework using Chain of Thought (CoT) and Group Relative Policy Optimization (GRPO) mechanisms, enabling anomaly detection directly from raw images without annotated masks. This approach generates interpretable step-by-step explanations for defect localization. Our method achieves state-of-the-art performance, outperforming prior approaches by 36% in accuracy on MVTec-AD and 16% on VisA. By eliminating mask dependency and reducing costs while providing explainable outputs, this work advances industrial anomaly detection and supports scalable quality control in manufacturing. Code to reproduce the experiment is available at https://github.com/LilaKen/LR-IAD.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
