用跨模态提示提升多类别异常定位精度，尤其擅长发现细微缺陷。
PromptMAD: Cross-Modal Prompting for Multi-Class Visual Anomaly Localization
- 通过文本提示融合正常与异常的语义信息，增强视觉重建
- 在MVTec-AD上达到98.35%的平均AUC和66.54%的AP
- 适合处理缺陷隐蔽、类别多样的工业检测场景
多类别视觉异常检测面临物体类别多样、异常样本稀少及伪装缺陷等问题。本文提出PromptMAD，一种基于跨模态提示的无监督异常检测与定位框架，通过视觉-语言对齐引入语义引导。利用CLIP编码的文本提示描述正常与异常类特定特征，丰富视觉重建的语义上下文，提升对细微纹理异常的检测能力。为缓解像素级类别不平衡问题，引入Focal损失函数，强化难检异常区域的学习。架构包含融合多尺度卷积特征与Transformer空间注意力的监督分割器，结合扩散迭代优化，生成高分辨率精准异常图。在MVTec-AD数据集上，本方法实现98.35%的平均AUC与66.54%的AP，性能达当前最优，且在各类别间保持高效。
原文摘要 · Abstract (English)
Visual anomaly detection in multi-class settings poses significant challenges due to the diversity of object categories, the scarcity of anomalous examples, and the presence of camouflaged defects. In this paper, we propose PromptMAD, a cross-modal prompting framework for unsupervised visual anomaly detection and localization that integrates semantic guidance through vision-language alignment. By leveraging CLIP-encoded text prompts describing both normal and anomalous class-specific characteristics, our method enriches visual reconstruction with semantic context, improving the detection of subtle and textural anomalies. To further address the challenge of class imbalance at the pixel level, we incorporate Focal loss function, which emphasizes hard-to-detect anomalous regions during training. Our architecture also includes a supervised segmentor that fuses multi-scale convolutional features with Transformer-based spatial attention and diffusion iterative refinement, yielding precise and high-resolution anomaly maps. Extensive experiments on the MVTec-AD dataset demonstrate that our method achieves state-of-the-art pixel-level performance, improving mean AUC to 98.35% and AP to 66.54%, while maintaining efficiency across diverse categories.
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。