CASIA OpenIR  > 精密感知与控制研究中心  > 精密感知与控制
Unsupervised Anomaly Detection for Surface Defects with Dual-Siamese Network
Tao X(陶显)1; Da-Peng Zhang1; Ma WZ(马文治)1; Hou ZX(侯占新)2; Lu ZF(逯正峰)2; Chandranath Adak3
Source PublicationIEEE Transactions on Industrial Informatics
ISSN1551-3203
2022-01
Volume1Issue:1Pages:1-11
Corresponding AuthorTao, Xian(taoxian2013@ia.ac.cn) ; Adak, Chandranath(adak32@gmial.com)
Subtype长文
Abstract

Unsupervised anomaly detection in real industrial scenarios is challenging since the small amount of defect-free images contain limited discriminative information, and anomaly defects are unpredictable. In this paper, a dual-siamese network is designed to simultaneously detect and locate anomalies in images. It first uses a pre-trained convolutional neural network (CNN)-based siamese architecture to embed discriminative features of normal samples and synthetic defective samples. A dense feature fusion (DFF) module is employed to obtain the dense feature representation of dual input. The following siamese network of perceptual defects is proposed to reconstruct and restore the dual-dense features of the previous stage. Compared to the existing methods that only employ a single residual map, the restoration of dense feature maps is proposed to locate the anomalies better. The experimental results on the MVTec AD dataset demonstrate that our method achieves state-of-the-art inspection accuracy and has potential for industrial application.

Keyword缺陷检测
MOST Discipline Catalogue工学
DOI10.1109/TII.2022.3142326
URL查看原文
Indexed BySCIE
Language英语
Funding ProjectBeijing Municipal Natural Science Foundation of China[4212044] ; National Natural Science Foundation of China[62066004] ; National Natural Science Foundation of China[62073317] ; Science Challenge[TZ2018006-0204-02]
Funding OrganizationBeijing Municipal Natural Science Foundation of China ; National Natural Science Foundation of China ; Science Challenge
WOS Research AreaAutomation & Control Systems ; Computer Science ; Engineering
WOS SubjectAutomation & Control Systems ; Computer Science, Interdisciplinary Applications ; Engineering, Industrial
WOS IDWOS:000856145200034
PublisherIEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
Sub direction classification目标检测、跟踪与识别
Citation statistics
Document Type期刊论文
Identifierhttp://ir.ia.ac.cn/handle/173211/47200
Collection精密感知与控制研究中心_精密感知与控制
Corresponding AuthorTao X(陶显)
Affiliation1.Institute of Automation, Chinese Academy of Sciences
2.China University of Mining and Technology - Beijing
3.Indian Institute of Information Technology
First Author AffilicationInstitute of Automation, Chinese Academy of Sciences
Corresponding Author AffilicationInstitute of Automation, Chinese Academy of Sciences
Recommended Citation
GB/T 7714
Tao X,Da-Peng Zhang,Ma WZ,et al. Unsupervised Anomaly Detection for Surface Defects with Dual-Siamese Network[J]. IEEE Transactions on Industrial Informatics,2022,1(1):1-11.
APA Tao X,Da-Peng Zhang,Ma WZ,Hou ZX,Lu ZF,&Chandranath Adak.(2022).Unsupervised Anomaly Detection for Surface Defects with Dual-Siamese Network.IEEE Transactions on Industrial Informatics,1(1),1-11.
MLA Tao X,et al."Unsupervised Anomaly Detection for Surface Defects with Dual-Siamese Network".IEEE Transactions on Industrial Informatics 1.1(2022):1-11.
Files in This Item: Download All
File Name/Size DocType Version Access License
Unsupervised_Anomaly(8384KB)期刊论文作者接受稿开放获取CC BY-NC-SAView Download
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Tao X(陶显)]'s Articles
[Da-Peng Zhang]'s Articles
[Ma WZ(马文治)]'s Articles
Baidu academic
Similar articles in Baidu academic
[Tao X(陶显)]'s Articles
[Da-Peng Zhang]'s Articles
[Ma WZ(马文治)]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Tao X(陶显)]'s Articles
[Da-Peng Zhang]'s Articles
[Ma WZ(马文治)]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: Unsupervised_Anomaly_Detection_for_Surface_Defects_with_Dual-Siamese_Network.pdf
Format: Adobe PDF
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.