计算机科学

计算机科学 ›› 2023, Vol. 50 ›› Issue (11): 151-159.doi: 10.11896/jsjkx.221100023

• 计算机图形学&多媒体 • 上一篇    下一篇

基于图像重构与语义差异识别的表面异常检测

王尚尚, 金城   

  1. 复旦大学计算机科学技术学院 上海 200438
  • 收稿日期:2022-11-03 修回日期:2023-03-16 出版日期:2023-11-15 发布日期:2023-11-06
  • 通讯作者: 金城(jc@fudan.edu.cn)
  • 作者简介:(shangshangwang20@fudan.edu.cn)
  • 基金资助:
    国家重点研发计划(2019YFB2102800)

Surface Anomaly Detection Based on Image Reconstruction and Semantic Difference Discrimination

WANG Shangshang, JIN Cheng   

  1. School of Computer Science,Fudan University,Shanghai 200438,China
  • Received:2022-11-03 Revised:2023-03-16 Online:2023-11-15 Published:2023-11-06
  • About author:WANG Shangshang,born in 1997,postgraduate.His main research interests include unsupervised anomaly detection and localization.JIN Cheng,born in 1978,Ph.D,professor,Ph.D supervisor,is a member of China Computer Federation.His main research interests include computer vision and multimedia information retrieval.
  • Supported by:
    National Key Research and Development Program of China(2019YFB2102800).

PDF (PC)

1846

摘要： 基于图像重构的方法是表面异常检测中一类广泛使用的方法。该类方法仅期望模型较好地重构正常模式,并通过异常区域较大的重构误差来检测和定位异常。已有方法一方面易出现“泛化”过好的现象,异常区域也被高保真地重构了出来;另一方面仅在图像空间度量重构误差,并没有真正捕捉到原图和重构图之间的语义差异。为了解决上述问题,文中提出了由重构网络和识别网络组成的表面异常检测框架,其中重构网络嵌入了多尺度位置增强动态原型单元,强化了对正常模式的学习;识别网络进行了输入图和重构图的多尺度深度特征融合,从多个尺度利用了重构前后的语义差异信息,强化了对重构差异的识别。在MVTec数据集上,所提方法在异常检测任务上取得了99.5%的 AUROC,在异常定位任务上取得了98.5% 的AUROC,以及95.0%的RPO检测表现,与之前基于重构的表面异常检测方法相比取得了较大提升。

关键词: 图像重构, 表面异常检测, 多尺度位置增强动态原型单元, 语义差异识别

Abstract: Reconstruction-based methods are widely used for surface anomaly detection.These methods are expected to only reconstruct normal patterns well and detect and localize anomalies by the larger reconstruction error in anomalous areas.Previous methods either tend to “generalize” too well,resulting in high fidelity reconstruction of anomalies,or measure reconstruction differences in image space,which doesn’t really capture the semantic differences.To tackle these problems,this paper proposes a model consisting of a reconstruction network and a discrimination network.In the reconstruction network,we design a multiscale location-augmented dynamic prototype unit to reinforce the learning of normal patterns.In the discrimination network,we fuse the multiscale deep features of the input image and its anomaly-free reconstruction to utilize the multiscale semantic difference information before and after reconstruction,which reinforces the discrimination of semantic differences.On the MVTec dataset,our method reaches 99.5% AUROC in the detection task,and 98.5% AUROC,95.0% PRO in the location task,outperforms pre-vious reconstruction-based methods by a large margin.

Key words: Image reconstruction, Surface anomaly detection, Multiscale location-augmented dynamic prototype unit, Semantic difference discrimination

中图分类号: 

  • TP391
[1]BERGMANN P,FAUSER M,SATTLEGGER D,et al.MVTec AD — A Comprehensive Real-World Dataset for Unsupervised Anomaly Detection[C]//2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR).IEEE,2020.
[2]SCHLEGL T,SEEBCK P,WALDSTEIN S M,et al.Unsupervised Anomaly Detection with Generative Adversarial Networks to Guide Marker Discovery [C]//International Conference on Information Processing in Medical Imaging.Cham:Springer,2017:146-157.
[3]ZAVRTANIK V,KRISTAN M,SKOAJ D.DR{\AEM－A discriminatively trained reconstruction embedding for surface anomaly detection[C]//IEEE/CVF International Conference on Computer Vision.2021:8330- 8339.
[4]GONG D,LIU L,LE V,et al.Memorizing Normality to Detect Anomaly:Memory-Augmented Deep Autoencoder for Unsupervised Anomaly Detection[C]//2019 IEEE/CVF International Conference on Computer Vision(ICCV).IEEE,2020.
[5]PARK H,NOH J,HAM B.Learning Memory-guided Normality for Anomaly Detection[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR).IEEE,2020.
[6]ZAVRTANIK V,KRISTAN M,SKOAJ D.Reconstruction by inpainting for visual anomaly detection[J].Pattern Recognition,2021,112:107706
[7]MEI S,YANG H,YIN Z.An Unsupervised-Learning-Based Approach for Automated Defect Inspection on Textured Surfaces[J].IEEE Transactions on Instrumentation and Measurement,2018,67(6):1266-1277.
[8]PARK C,CHO M A,LEE M,et al.FastAno:Fast Anomaly Detection via Spatio-temporal Patch Transformation[C]//IEEE/CVF Winter Conference on Applications of Computer Vision.2022:2249-2259.
[9]SABOKROU M,KHALOOEI M,FATHY M,et al.Adversa-rially Learned One-Class Classifier for Novelty Detection[C]//IEEE/CVF Conference on Computer Vision & Pattern Recognition.IEEE,2018.
[10]BERGMANN P,LWE S,FAUSER M,et al.Improving Unsupervised Defect Segmentation by Applying Structural Similarity to Autoencoders[J].arXiv:1807.02011,2018.
[11]YAN X,ZHANG H,XU X,et al.Learning Semantic Context from Normal Samples for Unsupervised Anomaly Detection[C]//National Conference on Artificial Intelligence.2021.
[12]LI Z,LI N,JIANG K,et al.Superpixel Masking and Inpainting for Self-Supervised Anomaly Detection[C]//British Machine Vision Conference.2020.
[13]AKCAY S,ATAPOUR-ABARGHOUEI A,BRECKON T P.GANomaly:semi-supervised anomaly detection via adversarial training [C]//Asian Conference on Computer Vision.Cham:Springer,2018:622-637.
[14]AKAY S,ATAPOUR-ABARGHOUEI A,BRECKON T P.Skip-GANomaly:Skip Connected and Adversarially Trained Encoder-Decoder Anomaly Detection[C]//2019 International Joint Conference on Neural Networks(IJCNN).IEEE,2019.
[15]POURREZA M,MOHAMMADI B,KHAKI M,et al.G2D:Generate to Detect Anomaly[C]//Workshop on Applications of Computer Vision.IEEE,2021.
[16]SCHLEGL T,SEEBCK P,WALDSTEIN S M,et al.f-Ano-GAN:Fast Unsupervised Anomaly Detection with Generative Adversarial Networks[J].Medical Image Analysis,2019,54:30-44.
[17]COLLIN A S,VLEESCHOUWER C D.Improved anomaly detection by training an autoencoder with skip connections on images corrupted with Stain-shaped noise[C]//25th International Conference on Pattern Recognition(ICPR).IEEE,2021:7915-7922.
[18]ZHAO Z,LI B,DONG R,et al.A surface defect detection me-thod based on positive samples[C]//Pacific RIM International Conference on Artificial Intelligence.2018:473-481.
[19]HOU J,ZHANG Y,ZHONG Q,et al.Divide-and-Assemble:Learning Block-wise Memory for Unsupervised Anomaly Detection[C]//IEEE/CVF International Conference on Computer Vision.2021:8791-8800.
[20]LV H,CHEN C,CUI Z,et al.Learning Normal Dynamics inVideos with Meta Prototype Network[C]//Computer Vision and Pattern Recognition.2021.
[21]DENG J,DONG W,SOCHER R,et al.ImageNet:a Large-Scale Hierarchical Image Database[C]//2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition(CVPR 2009).IEEE,2009:248-255.
[22]RIPPEL O,MERTENS P,MERHOF D.Modeling the Distribution of Normal Data in Pre-Trained Deep Features for Anomaly Detection[C]//2020 25th International Conference on Pattern Recognition(ICPR).IEEE,2021:6726-6733.
[23]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.Springer,2021:475-489.
[24]RUDOLPH M,WANDT B,ROSENHAHN B.Same Same But DifferNet:Semi-Supervised Defect Detection with Normalizing Flows[C]//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision.2021:1907-1916.
[25]GUDOVSKIY D,ISHIZAKA S,KOZUKA K.CFLOW-AD:Real-Time Unsupervised Anomaly Detection with Localization via Conditional Normalizing Flows[C]//IEEE/CVF Winter Confe-rence on Applications of Computer Vision.2022:98-107.
[26]REZENDE D J,MOHAMED S.Variational Inference with Normalizing Flows[C]//International Conference on Machine Learning.JMLR.org,2015.
[27]BERGMANN P,FAUSER M,SATTLEGGER D,et al.Unin-formed Students:Student-Teacher Anomaly Detection with Discriminative Latent Embeddings[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2020:4183-4192.
[28]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.
[29]DENG H,LI X.Anomaly Detection via Reverse Distillationfrom One-Class Embedding[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2022:9737-9746.
[30]ROTH K,PEMULA L,ZEPEDA J,et al.Towards Total Recall in Industrial Anomaly Detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2022:14318-14328.
[31]COHEN N,HOSHEN Y.Sub-Image Anomaly Detection withDeep Pyramid Correspondences[J].arXiv:2005.02357,2020.
[32]REISS T,COHEN N,BERGMAN L,et al.PANDA:Adapting Pretrained Features for Anomaly Detection and Segmentation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2021:2806-2814.
[33]TSAI C,WU T,LAI S,et al.Multi-scale patch-based representation learning for image anomaly detection and segmentation[C]//IEEE/CVF Winter Conference on Applications of Computer Vision.2022:3065-3073.
[34]RACKI D,TOMAZEVIC D,SKOCAJ D.A Compact Convolu-tional Neural Network for Textured Surface Anomaly Detection[C]//2018 IEEE Winter Conference on Applications of Computer Vision(WACV).IEEE,2018:1331-1339.
[35]RONNEBERGER O,FISCHER P,BROX T.U-Net:Convolu-tional Networks for Biomedical Image Segmentation[C]//International Conference on Medical Image Computing and Compu-ter-Assisted Intervention.Springer International Publishing,2015.
[36]LIN T Y,GOYAL P,GIRSHICK R,et al.Focal Loss for Dense Object Detection[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2017:2980-2988.
[37]WIELER M,HAHN T.Weakly supervised learning for industrial optical inspection [C]//DAGM Symposium.2007.
[38]CIMPOI M,MAJI S,KOKKINOS I,et al.Describing Textures in the Wild[C]//IEEE Conference on Computer Vision and Pattern Recognition.2014:3606-3613.
[39]SHI Y,YANG J,QI Z.Unsupervised anomaly segmentation via deep feature reconstruction[J].Neurocomputing,2020,424(11):9-22.
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