通过感知图像块的动态编码分配，提升无监督缺陷检测精度。
Patch-aware Vector Quantized Codebook Learning for Unsupervised Visual Defect Detection
- 引入块感知的动态代码分配机制，优化空间表征
- 在MVTecAD等3个数据集上达到顶尖检测性能
- 适合工业视觉质检场景，尤其关注高精度缺陷识别
无监督视觉缺陷检测在工业应用中至关重要，需构建能捕捉正常数据特征并识别异常的表征空间。在表达能力与紧凑性之间取得平衡颇具挑战：过于表达性的空间可能导致效率低下和模式坍缩，影响检测准确率。本文提出一种基于增强型VQ-VAE框架的新方法，专为无监督缺陷检测优化。模型引入块感知的动态代码分配策略，实现上下文敏感的代码分配，以优化空间表征。该策略增强了正常与缺陷之间的区分能力，在推理阶段显著提升检测精度。在MVTecAD、BTAD和MTSD三个数据集上的实验表明，该方法达到当前最优性能。
原文摘要 · Abstract (English)
Unsupervised visual defect detection is critical in industrial applications, requiring a representation space that captures normal data features while detecting deviations. Achieving a balance between expressiveness and compactness is challenging; an overly expressive space risks inefficiency and mode collapse, impairing detection accuracy. We propose a novel approach using an enhanced VQ-VAE framework optimized for unsupervised defect detection. Our model introduces a patch-aware dynamic code assignment scheme, enabling context-sensitive code allocation to optimize spatial representation. This strategy enhances normal-defect distinction and improves detection accuracy during inference. Experiments on MVTecAD, BTAD, and MTSD datasets show our method achieves state-of-the-art performance.
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。