计算机科学

计算机科学 ›› 2024, Vol. 51 ›› Issue (11A): 240200069-7.doi: 10.11896/jsjkx.240200069

• 图像处理&多媒体技术 • 上一篇    下一篇

面向工业图像异常检测的非对称师生网络模型

孔森林1, 张辉2, 黄镇南3, 刘优武1, 陶岩1   

  1. 1 长沙理工大学电气与信息工程学院 长沙 410000
    2 湖南大学机器人学院 长沙 410000
    3 中国人民武装警察部队警官学院 成都 610213
  • 出版日期:2024-11-16 发布日期:2024-11-13
  • 通讯作者: 张辉(zhanghuihby@126.com)
  • 作者简介:(986735244@qq.com)
  • 基金资助:
    科技创新2030－“新一代人工智能”重大项目(2021ZD0114503);国家自然科学基金重大研究计划(92148204);国家自然科学基金(62027810);湖南省科技创新领军人才(2022RC3063);湖南省杰出青年科学基金项目(2021JJ10025);湖南省重点研发计划(2021GK4011,2022GK2011);长沙科技重大项目(KH2003026);中国高校产学研创新基金(2020HYA06006);湖南省研究生科研创新项目(CX20220923);长沙理工大学研究生科研创新项目(CXCLY2022088)

Asymmetric Teacher-Student Network Model for Industrial Image Anomaly Detection

KONG Senlin1, ZHANG Hui2, HUANG Zhennan3, LIU Youwu1, TAO Yan1   

  1. 1 School of Electrical & Information Engineering,Changsha University of Science and Technology,Changsha 410000,China
    2 School of Robotics,Hunan University,Changsha 410000,China
    3 Officers College of PAP,Chengdu 610213,China
  • Online:2024-11-16 Published:2024-11-13
  • About author:KONG Senlin,born in 1997,master.His main research interests include unsupervised learning and industrial image defect detection.
    ZHANG Hui,born in 1983,Ph.D,professor,Ph.D supervisor.His main research interests include image proces-sing and robot vision detection.
  • Supported by:
    Science and Technology Innovation 2030－“New Generation Artificial Intelligence” Major Project(2021ZD0114503),National Natural Science Foundation of China Major Research Program(92148204),National Natural Science Foundation of China(62027810),Leading Scientific and Technological Innovation Talents of Hunan Province(2022RC3063),Hunan Outstanding Young People Science Foundation Project(2021JJ10025),Hunan Key Research and Development Project(2021GK4011,2022GK2011),Changsha Key Science and Technology Project(KH2003026),China University Industry University Research Innovation Fund (2020HYA06006),Hunan Graduate Research Innovation Project(CX20220923) and Changsha University of Science and Technology Graduate Research Innovation Project (CXCLY20222088).

PDF (PC)

197

摘要： 工业图像异常检测是大规模工业制造中的关键组成部分。针对工业图像异常检测存在的异常样本标注难度大、异常区域先验信息获取困难等问题,提出了一种基于非对称师生网络的无监督图像异常检测模型。首先,针对高相似结构师生网络导致的过模仿映射问题,设计了非对称师生网络,通过向学生网络残差块中引入上下文Transformer模块,为师生网络添加结构差异性,阻止学生网络过模仿教师网络的映射。其次,为了增强师生网络之间的泛化性差异,在教师网络中引入移动平均归一化层,以提高检测性能。最后,引入多尺度异常图融合机制,通过融合不同尺度的异常分数图,以更好地检测不同大小的异常。在MVTec AD公共数据集上进行了相关实验,实验结果中图像级别AUROC达到95.7%,像素级别AUROC达到97.4%,验证了该方法的可行性和有效性。

关键词: 异常检测, 知识蒸馏, Transformer, 无监督学习, 多尺度特征

Abstract: Industrial image anomaly detection is a critical component in large-scale industrial manufacturing.Addressing challenges such as difficulty in annotating anomalous samples and obtaining prior information about anomalous regions in industrial image anomaly detection,a model based on asymmetric teacher-student networks for unsupervised image anomaly detection is proposed.Firstly,to tackle the problem of over-imitation mapping caused by high similarity in structure between teacher and student networks,an asymmetric teacher-student network is designed.Contextual Transformer modules are introduced into the residual blocks of the student network to add structural diversity to the teacher-student networks,preventing the student network from over-imitating the mapping of the teacher network.Secondly,to enhance the generalization difference between teacher and student networks,a moving average normalization layer is introduced into the teacher network to improve detection performance.Finally,a multi-scale abnormality map fusion mechanism is introduced to better detect anomalies of different sizes by fusing abnormality score maps of different scales.Experiments conducted on the MVTec AD public dataset show that the proposed method achieves an image-level AUROC of 95.7% and a pixel-level AUROC of 97.4%,verifying the feasibility and effectiveness of the approach.

Key words: Anomaly detection, Knowledge distillation, Transformer, Unsupervised learning, Multi-scale features

中图分类号: 

  • TP391
[1]PANG G,SHEN C,CAO L,et al.Deep learning for anomaly detection:a review[J].ACM computing surveys(CSUR),2021,54(2):1-38.
[2]LOWE D G.Distinctive image features from scale-invariant keypoints[J].International Journal of Computer Vision,2004,60(2):91-110.
[3]ARRIS C,STEPHENS M.A combined corner and edge detector[C]//Alvey Vision Conference.1988.
[4]WANG Z,WU F,HU Z.Msld:a robust descriptor for line matching[J].Pattern Recognition,2009,42(5):941-953.
[5]ZHANG L,KOCH R.An efficient and robust line segmentmatching approach based on lbd descriptor and pairwise geometric consistency[J].Journal of Visual Communication and Image Representation,2013,24(7):794-805.
[6]NISTÉR D,STEWÉNIUS H.Linear time maximally stable extremal regions[C]//Computer Vision-ECCV 2008:10th European Conference on Computer Vision,Marseille,France,Part II 10.Springer Berlin Heidelberg,2008:183-196.
[7]XU Y,MONASSE P,GÉRAUD T,et al.Tree-based morse regions:a topological approach to local feature detection[J].IEEE Transactions on Image Processing,2014,23(12):5612-5625.
[8]GIRSHICK R,DONAHUE J,DARRELL T,et al.Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2014:580-587.
[9]CZIMMERMANN T,CIUTI G,MILAZZO M,et al.Visual-based defect detection and classification approaches for industrial applications－A survey[J].Sensors,2020,20(5):1459.
[10]REN S,HE K,GIRSHICK R,et al.Faster r-cnn:towards real-time object detection with region proposal networks[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2017,39(6):1137-1149.
[11]YANG L,ZHONG J,ZHANG Y,et al.An improving faster-rcnn with multi-attention resnet for small target detection in intelligent autonomous transport with 6g [J].IEEE Transactions on Intelligent Transportation Systems,2023,24(7):7717-7725.
[12]REDMON J,DIVVALA S,GIRSHICK R,et al.You only look once:unified,real-time object detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2016:779-788.
[13]JIANG P,ERGU D,LIU F,et al.A review of yolo algorithm developments[J].Procedia Computer Science,2022,199:1066-1073.
[14]LIU W,ANGUELOV D,ERHAN D,et al.Ssd:single shotmultibox detector[C]//Computer Vision-ECCV 2016:14th European Conference,Amsterdam,The Netherlands,Part I 14.Springer International Publishing,2016:21-37.
[15]ZHENG W,TANG W,JIANG L,et al.Se-ssd:self-ensembling single-stage object detector from point cloud[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2021:14494-14503.
[16]SAMARIYA D,THAKKAR A.A comprehensive survey of anomaly detection algorithms[J].Annals of Data Science,2023,10(3):829-850.
[17]BERGMANN P,LWE S,FAUSER M,et al.Improving unsu-pervised defect segmentation by applying structural similarity to autoencoders[C]//14th International Conference on Computer Vision Theory and Applications.2019.
[18]POL A A,BERGER V,GERMAIN C,et al.Anomaly detection with conditional variational autoencoders[C]//2019 18th IEEE International Conference on Machine Learning And applications(ICMLA).IEEE,2019:1651-1657.
[19]SONG J,KONG K,PARK Y I,et al.AnoSeg:anomaly segmentation network using self-supervised learning [EB/OL].(2021-10-7) [2024-02-19].https://doi.org/10.48550/arXiv.2110.03396,2021.
[20]MADAN N,RISTEA N C,IONESCU R T,et al.Self-supervised masked convolutional transformer block for anomaly detection[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2023,46(1):525-542.
[21]WAN Q,GAO L,LI X,et al.Unsupervised image anomaly detection and segmentation based on pretrained feature mapping[J].IEEE Transactions on Industrial Informatics,2022,19(3):2330-2339.
[22]COHEN N,HOSHEN Y.Sub-image anomaly detection withdeep pyramid correspondences [EB/OL].(2020-05-05)[2024-02-19].https://doi.org/10.48550/arXiv.2005.02357,2020.
[23]GUDOVSKIY D,ISHIZAKA S,KOZUKA K.Cflow-ad:real-time unsupervised anomaly detection with localization via conditional normalizing flows[C]//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision(WACV).2022:98-107.
[24]SALEHI M,SADJADI N,BASELIZADEH S,et al.Multiresolution knowledge distillation for anomaly detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2021:14902-14912.
[25]DEFARD T,SETKOV A,LOESCH A,et al.Padim:a patch distribution modeling framework for anomaly detection and localization[C]//International Conference on Pattern Recognition.Cham:Springer International Publishing,2021:475-489.
[26]HE K,ZHANG X,REN S,et al.Deep residual learning forimage recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2016:770-778.
[27]LI Y,YAO T,PAN Y,et al.Contextual transformer networks for visual recognition[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2022,45(2):1489-1500.
[28]CAI Z,RAVICHANDRAN A,MAJI S,et al.Exponential moving average normalization for self-supervised and semi-supervised learning[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2021:194-203.
[29]BERGMANN P,FAUSER M,SATTLEGGER D,et al.Mvtec ad－a comprehensive real-world dataset for unsupervised anomaly detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2019:9592-9600.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发