Journal of Computer Applications ›› 2024, Vol. 44 ›› Issue (11): 3587-3594.DOI: 10.11772/j.issn.1001-9081.2023111690

• Multimedia computing and computer simulation • Previous Articles     Next Articles

Pixel-level unsupervised industrial anomaly detection based on multi-scale memory bank

Yongjiang LIU1,2, Bin CHEN2,3,4()   

  1. 1.Chengdu Institute of Computer Application,Chinese Academy of Sciences,Chengdu Sichuan 610041,China
    2.University of Chinese Academy of Sciences,Beijing 100049,China
    3.International Institute for Artificial Intelligence,Harbin Institute of Technology (Shenzhen),Shenzhen Guangdong 518055,China
    4.Chongqing Research Institute of Harbin Institute of Technology,Chongqing 401151,China
  • Received:2023-12-08 Revised:2024-03-01 Accepted:2024-03-07 Online:2024-03-12 Published:2024-11-10
  • Contact: Bin CHEN
  • About author:LIU Yongjiang, born in 1996, M. S. candidate. His research interests include object detection, anomaly detection.

基于多尺度记忆库的像素级无监督工业异常检测

刘永江1,2, 陈斌2,3,4()   

  1. 1.中国科学院 成都计算机应用研究所,成都 610041
    2.中国科学院大学,北京 100049
    3.哈尔滨工业大学(深圳) 国际人工智能研究院,广东 深圳 518055
    4.哈尔滨工业大学 重庆研究院,重庆 401151
  • 通讯作者: 陈斌
  • 作者简介:刘永江(1996—),男,江西九江人,硕士研究生,主要研究方向:目标检测、异常检测

Abstract:

Unsupervised anomaly detection methods based on feature embedding often use patch-level features to localize anomalies. Patch-level features are competitive in image-level anomaly detection tasks, but suffer from insufficient accuracy in pixel-level localization. To address this issue, MemAD, a pixel-level anomaly detection method composed of a multi-scale memory bank and a segmentation network, was proposed. Firstly, a pre-trained feature extraction network was used to extract features from normal samples in the training set, thereby constructing a normal sample feature memory bank at three scales. Then, during the training of the segmentation network, difference features between simulated pseudo-anomaly sample features and the nearest normal sample features in the memory bank were calculated, thereby further guiding the segmentation network to learn how to locate anomalous pixels. Experimental results show that MemAD achieves image-level and pixel-level AUC (Area Under the Receiver Operating Characteristic curve) of 0.980 and 0.974 respectively on MVTec AD (MVTec Anomaly Detection) dataset, outperforming most existing methods and confirming the accuracy of the proposed method in pixel-level anomaly localization.

Key words: computer vision, unsupervised anomaly detection, feature embedding, memory bank, semantic segmentation

摘要:

基于特征嵌入的无监督异常检测方法通常使用patch级特征定位异常。patch级特征在图像级异常检测任务上具有竞争力,但在像素级定位方面存在精度不足的问题。为解决这一问题,提出一种由多尺度记忆库与分割网络组成的像素级异常检测方法MemAD。首先,通过预训练的特征提取网络对训练集中的正常样本进行特征提取,构建3个尺度下的正样本特征记忆库;其次,在训练分割网络时,计算模拟的伪异常样本特征与记忆库中距离最近的正样本特征的差特征,进一步引导分割网络学习如何定位异常像素。实验结果表明,MemAD在MVTec AD (MVTec Anomaly Detection)数据集上的图像级和像素级接受者操作特征曲线下面积(AUC)分别达到了98.0%和97.4%,优于大多数的现有方法,验证了它在像素级异常定位中的准确性。

关键词: 计算机视觉, 无监督异常检测, 特征嵌入, 记忆库, 语义分割

CLC Number: