Reducing Anomaly Detection in Images to Detection in Noise

Axel DAVY; Thibaud EHRET; Jean-Michel MOREL; Mauricio DELBRACIO

doi:10.1109/ICIP.2018.8451059

Reducing Anomaly Detection in Images to Detection in Noise

Axel DAVY
, Thibaud EHRET
, Jean-Michel MOREL
, Mauricio DELBRACIO

Research output: Book Chapters | Papers in Conference Proceedings › Conference paper (refereed) › Research › peer-review

19 Citations (Scopus)

Abstract

Anomaly detectors address the difficult problem of detecting automatically exceptions in an arbitrary background image. Detection methods have been proposed by the thousands because each problem requires a different background model. By analyzing the existing approaches, we show that the problem can be reduced to detecting anomalies in residual images (extracted from the target image) in which noise and anomalies prevail. Hence, the general and impossible background modeling problem is replaced by simpler noise modeling, and allows the calculation of rigorous thresholds based on the a contrario detection theory. Our approach is therefore unsupervised and works on arbitrary images.

Original language	English
Title of host publication	2018 IEEE International Conference on Image Processing, ICIP 2018, Proceedings
Publisher	IEEE
Pages	1058-1062
Number of pages	5
ISBN (Electronic)	9781479970612
ISBN (Print)	9781479970629
DOIs	https://doi.org/10.1109/ICIP.2018.8451059
Publication status	Published - 2018
Externally published	Yes
Event	25th IEEE International Conference on Image Processing, ICIP 2018 - Megaron Athens International Conference Centre, Athens, Greece Duration: 7 Oct 2018 → 10 Oct 2018

Publication series

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

Conference

Conference	25th IEEE International Conference on Image Processing, ICIP 2018
Country/Territory	Greece
City	Athens
Period	7/10/18 → 10/10/18

Bibliographical note

Publisher Copyright:
© 2018 IEEE.

Funding

Work supported by IDEX Paris-Saclay IDI 2016, ANR-11-IDEX-0003-02, ONR grant N00014-17-1-2552, CNES MISS project, Agencia Nacional de Investigacion e Innovacion (ANII, Uruguay) grant FCE 1 2017 135458, DGA Astrid ANR-17-ASTR-0013-01, DGA ANR-16-DEFA-0004-01, Programme ECOS Sud - UdelaR - Paris Descartes U17E04, and MENRT

Keywords

Anomaly detection
Saliency
Self-similarity

Access to Document

10.1109/ICIP.2018.8451059

Cite this

@inproceedings{9937b0d55d104e3eba232a680485de66,

title = "Reducing Anomaly Detection in Images to Detection in Noise",

abstract = "Anomaly detectors address the difficult problem of detecting automatically exceptions in an arbitrary background image. Detection methods have been proposed by the thousands because each problem requires a different background model. By analyzing the existing approaches, we show that the problem can be reduced to detecting anomalies in residual images (extracted from the target image) in which noise and anomalies prevail. Hence, the general and impossible background modeling problem is replaced by simpler noise modeling, and allows the calculation of rigorous thresholds based on the a contrario detection theory. Our approach is therefore unsupervised and works on arbitrary images.",

keywords = "Anomaly detection, Saliency, Self-similarity",

author = "Axel DAVY and Thibaud EHRET and Jean-Michel MOREL and Mauricio DELBRACIO",

note = "Publisher Copyright: {\textcopyright} 2018 IEEE.; 25th IEEE International Conference on Image Processing, ICIP 2018 ; Conference date: 07-10-2018 Through 10-10-2018",

year = "2018",

doi = "10.1109/ICIP.2018.8451059",

language = "English",

isbn = "9781479970629",

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

publisher = "IEEE",

pages = "1058--1062",

booktitle = "2018 IEEE International Conference on Image Processing, ICIP 2018, Proceedings",

}

DAVY, A, EHRET, T, MOREL, J-M & DELBRACIO, M 2018, Reducing Anomaly Detection in Images to Detection in Noise. in 2018 IEEE International Conference on Image Processing, ICIP 2018, Proceedings., 8451059, Proceedings - International Conference on Image Processing, ICIP, IEEE, pp. 1058-1062, 25th IEEE International Conference on Image Processing, ICIP 2018, Athens, Greece, 7/10/18. https://doi.org/10.1109/ICIP.2018.8451059

Reducing Anomaly Detection in Images to Detection in Noise. / DAVY, Axel; EHRET, Thibaud; MOREL, Jean-Michel et al.
2018 IEEE International Conference on Image Processing, ICIP 2018, Proceedings. IEEE, 2018. p. 1058-1062 8451059 (Proceedings - International Conference on Image Processing, ICIP).

Research output: Book Chapters | Papers in Conference Proceedings › Conference paper (refereed) › Research › peer-review

TY - GEN

T1 - Reducing Anomaly Detection in Images to Detection in Noise

AU - DAVY, Axel

AU - EHRET, Thibaud

AU - MOREL, Jean-Michel

AU - DELBRACIO, Mauricio

PY - 2018

Y1 - 2018

N2 - Anomaly detectors address the difficult problem of detecting automatically exceptions in an arbitrary background image. Detection methods have been proposed by the thousands because each problem requires a different background model. By analyzing the existing approaches, we show that the problem can be reduced to detecting anomalies in residual images (extracted from the target image) in which noise and anomalies prevail. Hence, the general and impossible background modeling problem is replaced by simpler noise modeling, and allows the calculation of rigorous thresholds based on the a contrario detection theory. Our approach is therefore unsupervised and works on arbitrary images.

AB - Anomaly detectors address the difficult problem of detecting automatically exceptions in an arbitrary background image. Detection methods have been proposed by the thousands because each problem requires a different background model. By analyzing the existing approaches, we show that the problem can be reduced to detecting anomalies in residual images (extracted from the target image) in which noise and anomalies prevail. Hence, the general and impossible background modeling problem is replaced by simpler noise modeling, and allows the calculation of rigorous thresholds based on the a contrario detection theory. Our approach is therefore unsupervised and works on arbitrary images.

KW - Anomaly detection

KW - Saliency

KW - Self-similarity

UR - https://www.scopus.com/pages/publications/85062900578

U2 - 10.1109/ICIP.2018.8451059

DO - 10.1109/ICIP.2018.8451059

M3 - Conference paper (refereed)

AN - SCOPUS:85062900578

SN - 9781479970629

T3 - Proceedings - International Conference on Image Processing, ICIP

SP - 1058

EP - 1062

BT - 2018 IEEE International Conference on Image Processing, ICIP 2018, Proceedings

PB - IEEE

T2 - 25th IEEE International Conference on Image Processing, ICIP 2018

Y2 - 7 October 2018 through 10 October 2018

ER -

Reducing Anomaly Detection in Images to Detection in Noise

Abstract

Publication series

Conference

Bibliographical note

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this