Unsupervised Variability Normalization For Anomaly Detection

Aitor Artola; Yannis Kolodziej; Jean-Michel Morel; Thibaud Ehret

doi:10.1109/ICIP42928.2021.9506742

Unsupervised Variability Normalization For Anomaly Detection

Aitor Artola
, Yannis Kolodziej
, Jean-Michel Morel
, Thibaud Ehret

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

1 Citation (Scopus)

Abstract

Anomaly detectors are necessary to automatize industrial quality control. However, crafting such detectors is difficult due to the complexity and variability of the object even when working only with rigid objects. We show that adding a deep learning normalization step as a preprocessing step to model based detectors allows for better and more robust detections. This self-supervised normalization neural network is trained on non-anomalous data only. The proposed preprocessing method, followed by an automatic detector, achieves state-of-the-art results on rigid objects from the MvTec dataset. © 2021 IEEE.

Original language	English
Title of host publication	2021 IEEE International Conference on Image Processing
Subtitle of host publication	PROCEEDINGS
Publisher	IEEE
Pages	989-993
ISBN (Electronic)	978-1-6654-3102-6
ISBN (Print)	9781665441155
DOIs	https://doi.org/10.1109/ICIP42928.2021.9506742
Publication status	Published - 2021
Externally published	Yes
Event	28th IEEE International Conference on Image Processing (ICIP 2021) - Denaʼina Civic and Convention Center, Anchorage, United States Duration: 19 Sept 2021 → 22 Sept 2021 https://www.2021.ieeeicip.org/ https://2021.ieeeicip.org/Papers/AcceptedPapers.asp

Publication series

Name	Proceedings - International Conference on Image Processing, ICIP
ISSN (Print)	1522-4880
ISSN (Electronic)	2381-8549

Conference

Conference	28th IEEE International Conference on Image Processing (ICIP 2021)
Abbreviated title	IEEE ICIP 2021
Place	United States
City	Anchorage
Period	19/09/21 → 22/09/21
Internet address	https://www.2021.ieeeicip.org/ https://2021.ieeeicip.org/Papers/AcceptedPapers.asp

Research Keywords

Anomaly detection
Deep-learning
Denoising
Self-similarity

Access Output

10.1109/ICIP42928.2021.9506742

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

@inproceedings{b0c982fd9d4044c7b6946d062ce17475,

title = "Unsupervised Variability Normalization For Anomaly Detection",

abstract = "Anomaly detectors are necessary to automatize industrial quality control. However, crafting such detectors is difficult due to the complexity and variability of the object even when working only with rigid objects. We show that adding a deep learning normalization step as a preprocessing step to model based detectors allows for better and more robust detections. This self-supervised normalization neural network is trained on non-anomalous data only. The proposed preprocessing method, followed by an automatic detector, achieves state-of-the-art results on rigid objects from the MvTec dataset. {\textcopyright} 2021 IEEE.",

keywords = "Anomaly detection, Deep-learning, Denoising, Self-similarity",

author = "Aitor Artola and Yannis Kolodziej and Jean-Michel Morel and Thibaud Ehret",

year = "2021",

doi = "10.1109/ICIP42928.2021.9506742",

language = "English",

isbn = "9781665441155",

series = "Proceedings - International Conference on Image Processing, ICIP",

publisher = "IEEE",

pages = "989--993",

booktitle = "2021 IEEE International Conference on Image Processing",

address = "United States",

note = "28th IEEE International Conference on Image Processing (ICIP 2021), IEEE ICIP 2021 ; Conference date: 19-09-2021 Through 22-09-2021",

url = "https://www.2021.ieeeicip.org/, https://2021.ieeeicip.org/Papers/AcceptedPapers.asp",

}

Artola, A, Kolodziej, Y, Morel, J-M & Ehret, T 2021, Unsupervised Variability Normalization For Anomaly Detection. in 2021 IEEE International Conference on Image Processing: PROCEEDINGS. Proceedings - International Conference on Image Processing, ICIP, IEEE, pp. 989-993, 28th IEEE International Conference on Image Processing (ICIP 2021), Anchorage, Alaska, United States, 19/09/21. https://doi.org/10.1109/ICIP42928.2021.9506742

Unsupervised Variability Normalization For Anomaly Detection. / Artola, Aitor; Kolodziej, Yannis; Morel, Jean-Michel et al.
2021 IEEE International Conference on Image Processing: PROCEEDINGS. IEEE, 2021. p. 989-993 (Proceedings - International Conference on Image Processing, ICIP).

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

TY - GEN

T1 - Unsupervised Variability Normalization For Anomaly Detection

AU - Artola, Aitor

AU - Kolodziej, Yannis

AU - Morel, Jean-Michel

AU - Ehret, Thibaud

PY - 2021

Y1 - 2021

N2 - Anomaly detectors are necessary to automatize industrial quality control. However, crafting such detectors is difficult due to the complexity and variability of the object even when working only with rigid objects. We show that adding a deep learning normalization step as a preprocessing step to model based detectors allows for better and more robust detections. This self-supervised normalization neural network is trained on non-anomalous data only. The proposed preprocessing method, followed by an automatic detector, achieves state-of-the-art results on rigid objects from the MvTec dataset. © 2021 IEEE.

AB - Anomaly detectors are necessary to automatize industrial quality control. However, crafting such detectors is difficult due to the complexity and variability of the object even when working only with rigid objects. We show that adding a deep learning normalization step as a preprocessing step to model based detectors allows for better and more robust detections. This self-supervised normalization neural network is trained on non-anomalous data only. The proposed preprocessing method, followed by an automatic detector, achieves state-of-the-art results on rigid objects from the MvTec dataset. © 2021 IEEE.

KW - Anomaly detection

KW - Deep-learning

KW - Denoising

KW - Self-similarity

UR - https://www.scopus.com/pages/publications/85125595015

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85125595015&origin=recordpage

U2 - 10.1109/ICIP42928.2021.9506742

DO - 10.1109/ICIP42928.2021.9506742

M3 - RGC 32 - Refereed conference paper (with host publication)

SN - 9781665441155

T3 - Proceedings - International Conference on Image Processing, ICIP

SP - 989

EP - 993

BT - 2021 IEEE International Conference on Image Processing

PB - IEEE

T2 - 28th IEEE International Conference on Image Processing (ICIP 2021)

Y2 - 19 September 2021 through 22 September 2021

ER -

Unsupervised Variability Normalization For Anomaly Detection

Abstract

Publication series

Conference

Research Keywords

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this