提出无监督组件感知异常生成方法，提升工业异常检测真实性和效果。
Component-aware Unsupervised Logical Anomaly Generation for Industrial Anomaly Detection
- 基于组件解耦学习，分步生成符合逻辑的异常图像
- 在MVTecLOCO上达91.2%的AUROC，优于现有方法
- 无需真实异常样本，适合工业场景快速部署
异常检测在工业制造中对保障产品质量和提升自动化效率至关重要。由于异常样本稀缺，传统检测方法受限，异常生成成为扩充数据集的关键。然而，现有生成模型常产生不真实的异常，导致误报率上升，或需依赖真实异常样本进行训练。本文将异常生成视为组合问题，提出无监督的组件感知框架ComGEN，通过多组件学习策略解耦视觉成分，并引入注意力引导的残差映射与多尺度迭代匹配参考进行生成编辑。该方法利用解耦后的文本-组件对揭示内在逻辑约束，实现更合理的异常生成。在MVTecLOCO数据集上的实验验证了其有效性，取得91.2%的最高AUROC。进一步在柴油发动机真实场景及广泛应用的MVTecAD数据集上测试表明，集成ComGEN生成的模拟异常后，自动化生产流程性能显著提升。
原文摘要 · Abstract (English)
Anomaly detection is critical in industrial manufacturing for ensuring product quality and improving efficiency in automated processes. The scarcity of anomalous samples limits traditional detection methods, making anomaly generation essential for expanding the data repository. However, recent generative models often produce unrealistic anomalies increasing false positives, or require real-world anomaly samples for training. In this work, we treat anomaly generation as a compositional problem and propose ComGEN, a component-aware and unsupervised framework that addresses the gap in logical anomaly generation. Our method comprises a multi-component learning strategy to disentangle visual components, followed by subsequent generation editing procedures. Disentangled text-to-component pairs, revealing intrinsic logical constraints, conduct attention-guided residual mapping and model training with iteratively matched references across multiple scales. Experiments on the MVTecLOCO dataset confirm the efficacy of ComGEN, achieving the best AUROC score of 91.2%. Additional experiments on the real-world scenario of Diesel Engine and widely-used MVTecAD dataset demonstrate significant performance improvements when integrating simulated anomalies generated by ComGEN into automated production workflows.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。