CCTV Image Sequence Generation and Modeling Method for Video Anomaly Detection Using Generative Adversarial Network

Wonsup Shin; Sung Bae Cho

doi:10.1007/978-3-030-03493-1_48

CCTV Image Sequence Generation and Modeling Method for Video Anomaly Detection Using Generative Adversarial Network

Wonsup Shin, Sung Bae Cho

College of Computing

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

5 Citations (Scopus)

Abstract

Video anomaly detection is one of the most attractive problem in various fields likes computer vision. In this paper, we propose a VAD classifier modeling method that learns in a supervised learning manner. The basic idea is to solve the problem of labeled data shortage through transfer learning. The key idea is to create an underlying model of transfer learning through the GAN of discriminator. We solved this problem by proposing a GAN model consisting of a generator that generates video sequences and a discriminator that follows LRCN structure. As a result of the experiment, The VAD classifier learned through GAN-based transfer learning obtained higher accuracy and recall than the pure LRCN classifier and other machine learning methods. Additionally, we demonstrated that the generator be able to stably generate the image similar to the actual data as the learning progressed. To the best of our knowledge, this paper is the first case to solve the VAD problem using the GAN model and the supervised learning manner.

Original language	English
Title of host publication	Intelligent Data Engineering and Automated Learning – IDEAL 2018 - 19th International Conference, Proceedings
Editors	Hujun Yin, Paulo Novais, David Camacho, Antonio J. Tallón-Ballesteros
Publisher	Springer Verlag
Pages	457-467
Number of pages	11
ISBN (Print)	9783030034924
DOIs	https://doi.org/10.1007/978-3-030-03493-1_48
Publication status	Published - 2018
Event	19th International Conference on Intelligent Data Engineering and Automated Learning, IDEAL 2018 - Madrid, Spain Duration: 2018 Nov 21 → 2018 Nov 23

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11314 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	19th International Conference on Intelligent Data Engineering and Automated Learning, IDEAL 2018
Country/Territory	Spain
City	Madrid
Period	18/11/21 → 18/11/23

Bibliographical note

Funding Information:
This work was supported by the ICT R&D program of MSIP/IITP. [2017-0-00306, Development of Multimodal Sensor-based Intelligent Systems for Outdoor Surveillance Robots].

Funding Information:
Acknowledgement. This work was supported by the ICT R&D program of MSIP/IITP. [2017-0-00306, Development of Multimodal Sensor-based Intelligent Systems for Outdoor Surveillance Robots].

Publisher Copyright:
© 2018, Springer Nature Switzerland AG.

All Science Journal Classification (ASJC) codes

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-030-03493-1_48

Cite this

Shin, W., & Cho, S. B. (2018). CCTV Image Sequence Generation and Modeling Method for Video Anomaly Detection Using Generative Adversarial Network. In H. Yin, P. Novais, D. Camacho, & A. J. Tallón-Ballesteros (Eds.), Intelligent Data Engineering and Automated Learning – IDEAL 2018 - 19th International Conference, Proceedings (pp. 457-467). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11314 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-03493-1_48

Shin, Wonsup ; Cho, Sung Bae. / CCTV Image Sequence Generation and Modeling Method for Video Anomaly Detection Using Generative Adversarial Network. Intelligent Data Engineering and Automated Learning – IDEAL 2018 - 19th International Conference, Proceedings. editor / Hujun Yin ; Paulo Novais ; David Camacho ; Antonio J. Tallón-Ballesteros. Springer Verlag, 2018. pp. 457-467 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{0fec5ad836d84426b662e07f45d435b4,

title = "CCTV Image Sequence Generation and Modeling Method for Video Anomaly Detection Using Generative Adversarial Network",

abstract = "Video anomaly detection is one of the most attractive problem in various fields likes computer vision. In this paper, we propose a VAD classifier modeling method that learns in a supervised learning manner. The basic idea is to solve the problem of labeled data shortage through transfer learning. The key idea is to create an underlying model of transfer learning through the GAN of discriminator. We solved this problem by proposing a GAN model consisting of a generator that generates video sequences and a discriminator that follows LRCN structure. As a result of the experiment, The VAD classifier learned through GAN-based transfer learning obtained higher accuracy and recall than the pure LRCN classifier and other machine learning methods. Additionally, we demonstrated that the generator be able to stably generate the image similar to the actual data as the learning progressed. To the best of our knowledge, this paper is the first case to solve the VAD problem using the GAN model and the supervised learning manner.",

author = "Wonsup Shin and Cho, {Sung Bae}",

note = "Funding Information: This work was supported by the ICT R&D program of MSIP/IITP. [2017-0-00306, Development of Multimodal Sensor-based Intelligent Systems for Outdoor Surveillance Robots]. Funding Information: Acknowledgement. This work was supported by the ICT R&D program of MSIP/IITP. [2017-0-00306, Development of Multimodal Sensor-based Intelligent Systems for Outdoor Surveillance Robots]. Publisher Copyright: {\textcopyright} 2018, Springer Nature Switzerland AG.; 19th International Conference on Intelligent Data Engineering and Automated Learning, IDEAL 2018 ; Conference date: 21-11-2018 Through 23-11-2018",

year = "2018",

doi = "10.1007/978-3-030-03493-1_48",

language = "English",

isbn = "9783030034924",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "457--467",

editor = "Hujun Yin and Paulo Novais and David Camacho and Tall{\'o}n-Ballesteros, {Antonio J.}",

booktitle = "Intelligent Data Engineering and Automated Learning – IDEAL 2018 - 19th International Conference, Proceedings",

address = "Germany",

}

Shin, W & Cho, SB 2018, CCTV Image Sequence Generation and Modeling Method for Video Anomaly Detection Using Generative Adversarial Network. in H Yin, P Novais, D Camacho & AJ Tallón-Ballesteros (eds), Intelligent Data Engineering and Automated Learning – IDEAL 2018 - 19th International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11314 LNCS, Springer Verlag, pp. 457-467, 19th International Conference on Intelligent Data Engineering and Automated Learning, IDEAL 2018, Madrid, Spain, 18/11/21. https://doi.org/10.1007/978-3-030-03493-1_48

CCTV Image Sequence Generation and Modeling Method for Video Anomaly Detection Using Generative Adversarial Network. / Shin, Wonsup; Cho, Sung Bae.
Intelligent Data Engineering and Automated Learning – IDEAL 2018 - 19th International Conference, Proceedings. ed. / Hujun Yin; Paulo Novais; David Camacho; Antonio J. Tallón-Ballesteros. Springer Verlag, 2018. p. 457-467 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11314 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - CCTV Image Sequence Generation and Modeling Method for Video Anomaly Detection Using Generative Adversarial Network

AU - Shin, Wonsup

AU - Cho, Sung Bae

N1 - Funding Information: This work was supported by the ICT R&D program of MSIP/IITP. [2017-0-00306, Development of Multimodal Sensor-based Intelligent Systems for Outdoor Surveillance Robots]. Funding Information: Acknowledgement. This work was supported by the ICT R&D program of MSIP/IITP. [2017-0-00306, Development of Multimodal Sensor-based Intelligent Systems for Outdoor Surveillance Robots]. Publisher Copyright: © 2018, Springer Nature Switzerland AG.

PY - 2018

Y1 - 2018

N2 - Video anomaly detection is one of the most attractive problem in various fields likes computer vision. In this paper, we propose a VAD classifier modeling method that learns in a supervised learning manner. The basic idea is to solve the problem of labeled data shortage through transfer learning. The key idea is to create an underlying model of transfer learning through the GAN of discriminator. We solved this problem by proposing a GAN model consisting of a generator that generates video sequences and a discriminator that follows LRCN structure. As a result of the experiment, The VAD classifier learned through GAN-based transfer learning obtained higher accuracy and recall than the pure LRCN classifier and other machine learning methods. Additionally, we demonstrated that the generator be able to stably generate the image similar to the actual data as the learning progressed. To the best of our knowledge, this paper is the first case to solve the VAD problem using the GAN model and the supervised learning manner.

AB - Video anomaly detection is one of the most attractive problem in various fields likes computer vision. In this paper, we propose a VAD classifier modeling method that learns in a supervised learning manner. The basic idea is to solve the problem of labeled data shortage through transfer learning. The key idea is to create an underlying model of transfer learning through the GAN of discriminator. We solved this problem by proposing a GAN model consisting of a generator that generates video sequences and a discriminator that follows LRCN structure. As a result of the experiment, The VAD classifier learned through GAN-based transfer learning obtained higher accuracy and recall than the pure LRCN classifier and other machine learning methods. Additionally, we demonstrated that the generator be able to stably generate the image similar to the actual data as the learning progressed. To the best of our knowledge, this paper is the first case to solve the VAD problem using the GAN model and the supervised learning manner.

UR - http://www.scopus.com/inward/record.url?scp=85057134660&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85057134660&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-03493-1_48

DO - 10.1007/978-3-030-03493-1_48

M3 - Conference contribution

AN - SCOPUS:85057134660

SN - 9783030034924

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 457

EP - 467

BT - Intelligent Data Engineering and Automated Learning – IDEAL 2018 - 19th International Conference, Proceedings

A2 - Yin, Hujun

A2 - Novais, Paulo

A2 - Camacho, David

A2 - Tallón-Ballesteros, Antonio J.

PB - Springer Verlag

T2 - 19th International Conference on Intelligent Data Engineering and Automated Learning, IDEAL 2018

Y2 - 21 November 2018 through 23 November 2018

ER -

Shin W, Cho SB. CCTV Image Sequence Generation and Modeling Method for Video Anomaly Detection Using Generative Adversarial Network. In Yin H, Novais P, Camacho D, Tallón-Ballesteros AJ, editors, Intelligent Data Engineering and Automated Learning – IDEAL 2018 - 19th International Conference, Proceedings. Springer Verlag. 2018. p. 457-467. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-03493-1_48

CCTV Image Sequence Generation and Modeling Method for Video Anomaly Detection Using Generative Adversarial Network

Abstract

Publication series

Other

Bibliographical note

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this