arXiv 2504.118952025-05-08cs.CV

搜索即所需：面向少样本异常检测

Search is All You Need for Few-shot Anomaly Detection

Academy for Engineering and Technology Fudan University（复旦大学工程与技术学院）；School of Biomedical Engineering Tsinghua University（清华大学生物医学工程学院）；School of Computer Science Fudan university（复旦大学计算机学院）；College of Design & Innovation Tongji University（同济大学设计与创新学院）；School of Physics and Mechanical & Electrical Engineering Hexi University（河西大学物理与机械电子工程学院）；Engineering Research Center of AI&Robotics, Ministry of Education（教育部人工智能与机器人工程研究中心）；Academy for Engineering&Technology Fudan university（复旦大学工程与技术学院）

Qishan Wang, Jia Guo, Shuyong Gao, Haofen Wang, Li Xiong, Junjie Hu, Hanqi Guo, Wenqiang Zhang

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出VisionAD，通过最近邻搜索框架结合视觉基础模型、双重增强、多层特征集成和类感知记忆库，在少样本异常检测中以无需训练的方式超越现有最先进方法。

摘要：

少样本异常检测（FSAD）已成为工业检测中一项关键但具有挑战性的任务，其中正常分布建模必须仅用少量正常图像完成。现有方法通常采用结合语言与视觉模态的多模态基础模型进行提示引导的异常检测，但这些方法往往需要复杂的提示工程和大量人工调参。在本文中，我们证明了一个直接的最近邻搜索框架能够在单类和多类FSAD场景中超越最先进的性能。我们提出的方法VisionAD由四个简单但必不可少的组件构成：（1）可扩展的视觉基础模型，用于提取通用且具有判别性的特征；（2）双重增强策略——支持增强以提升特征匹配适应性，查询增强以解决单视图预测的疏漏；（3）多层特征集成，以最小的计算开销同时捕获低频全局上下文和高频局部细节；（4）类感知视觉记忆库，实现高效的一对多多类检测。在MVTec-AD、VisA和Real-IAD基准上的广泛评估证明了VisionAD的卓越性能。仅使用1张正常图像作为支持，我们的方法分别取得了97.4%、94.8%和70.8%的显著图像级AUROC分数，以显著优势（+1.6%、+3.2%和+1.4%）超越了当前最先进的方法。VisionAD无需训练的特性和卓越的少样本能力使其对样本稀缺或获取成本高昂的实际应用特别有吸引力。代码可在https://github.com/Qiqigeww/VisionAD获取。

英文摘要：

Few-shot anomaly detection (FSAD) has emerged as a crucial yet challenging task in industrial inspection, where normal distribution modeling must be accomplished with only a few normal images. While existing approaches typically employ multi-modal foundation models combining language and vision modalities for prompt-guided anomaly detection, these methods often demand sophisticated prompt engineering and extensive manual tuning. In this paper, we demonstrate that a straightforward nearest-neighbor search framework can surpass state-of-the-art performance in both single-class and multi-class FSAD scenarios. Our proposed method, VisionAD, consists of four simple yet essential components: (1) scalable vision foundation models that extract universal and discriminative features; (2) dual augmentation strategies - support augmentation to enhance feature matching adaptability and query augmentation to address the oversights of single-view prediction; (3) multi-layer feature integration that captures both low-frequency global context and high-frequency local details with minimal computational overhead; and (4) a class-aware visual memory bank enabling efficient one-for-all multi-class detection. Extensive evaluations across MVTec-AD, VisA, and Real-IAD benchmarks demonstrate VisionAD's exceptional performance. Using only 1 normal images as support, our method achieves remarkable image-level AUROC scores of 97.4%, 94.8%, and 70.8% respectively, outperforming current state-of-the-art approaches by significant margins (+1.6%, +3.2%, and +1.4%). The training-free nature and superior few-shot capabilities of VisionAD make it particularly appealing for real-world applications where samples are scarce or expensive to obtain. Code is available at https://github.com/Qiqigeww/VisionAD.

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。