Deep Convolutional Generative Adversarial Networks for Flame Detection in Video

Süleyman Aslan; Uğur Güdükbay; B. Uğur Töreyin; A. Enis Çetin

doi:10.1007/978-3-030-63007-2_63

Deep Convolutional Generative Adversarial Networks for Flame Detection in Video

Süleyman Aslan, Uğur Güdükbay, B. Uğur Töreyin^*, A. Enis Çetin

^*Corresponding author for this work

Informatics Institute

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

9 Citations (Scopus)

Abstract

Real-time flame detection is crucial in video-based surveillance systems. We propose a vision-based method to detect flames using Deep Convolutional Generative Adversarial Neural Networks (DCGANs). Many existing supervised learning approaches using convolutional neural networks do not take temporal information into account and require a substantial amount of labeled data. To have a robust representation of sequences with and without flame, we propose a two-stage training of a DCGAN exploiting spatio-temporal flame evolution. Our training framework includes the regular training of a DCGAN with real spatio-temporal images, namely, temporal slice images, and noise vectors, and training the discriminator separately using the temporal flame images without the generator. Experimental results show that the proposed method effectively detects flame in video with negligible false-positive rates in real-time.

Original language	English
Title of host publication	Computational Collective Intelligence - 12th International Conference, ICCCI 2020, Proceedings
Editors	Ngoc Thanh Nguyen, Ngoc Thanh Nguyen, Bao Hung Hoang, Cong Phap Huynh, Dosam Hwang, Bogdan Trawinski, Gottfried Vossen
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	807-815
Number of pages	9
ISBN (Print)	9783030630065
DOIs	https://doi.org/10.1007/978-3-030-63007-2_63
Publication status	Published - 2020
Event	12th International Conference on Computational Collective Intelligence, ICCCI 2020 - Da Nang, Viet Nam Duration: 30 Nov 2020 → 3 Dec 2020

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	12496 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	12th International Conference on Computational Collective Intelligence, ICCCI 2020
Country/Territory	Viet Nam
City	Da Nang
Period	30/11/20 → 3/12/20

Bibliographical note

Publisher Copyright:
© 2020, Springer Nature Switzerland AG.

Funding

A. Enis C¸ etin’s research is partially funded by NSF with grant number 1739396 and NVIDIA Corporation. B. U˘gur Töreyin’s research is partially funded by TÜBİTAK 114E426, İTÜ BAP MGA-2017-40964 and MOA-2019-42321.

Funders	Funder number
TÜBİTAK	MOA-2019-42321, İTÜ BAP MGA-2017-40964, 114E426
National Science Foundation	1739396
NVIDIA

Keywords

Deep Convolutional Generative Adversarial Neural Network
Fire detection
Flame detection

Access to Document

10.1007/978-3-030-63007-2_63

Cite this

Aslan, S., Güdükbay, U., Töreyin, B. U., & Çetin, A. E. (2020). Deep Convolutional Generative Adversarial Networks for Flame Detection in Video. In N. T. Nguyen, N. T. Nguyen, B. H. Hoang, C. P. Huynh, D. Hwang, B. Trawinski, & G. Vossen (Eds.), Computational Collective Intelligence - 12th International Conference, ICCCI 2020, Proceedings (pp. 807-815). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12496 LNAI). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-63007-2_63

Aslan, Süleyman ; Güdükbay, Uğur ; Töreyin, B. Uğur et al. / Deep Convolutional Generative Adversarial Networks for Flame Detection in Video. Computational Collective Intelligence - 12th International Conference, ICCCI 2020, Proceedings. editor / Ngoc Thanh Nguyen ; Ngoc Thanh Nguyen ; Bao Hung Hoang ; Cong Phap Huynh ; Dosam Hwang ; Bogdan Trawinski ; Gottfried Vossen. Springer Science and Business Media Deutschland GmbH, 2020. pp. 807-815 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{4358d5b40b5b45f0ade3b62b7f64eb25,

title = "Deep Convolutional Generative Adversarial Networks for Flame Detection in Video",

abstract = "Real-time flame detection is crucial in video-based surveillance systems. We propose a vision-based method to detect flames using Deep Convolutional Generative Adversarial Neural Networks (DCGANs). Many existing supervised learning approaches using convolutional neural networks do not take temporal information into account and require a substantial amount of labeled data. To have a robust representation of sequences with and without flame, we propose a two-stage training of a DCGAN exploiting spatio-temporal flame evolution. Our training framework includes the regular training of a DCGAN with real spatio-temporal images, namely, temporal slice images, and noise vectors, and training the discriminator separately using the temporal flame images without the generator. Experimental results show that the proposed method effectively detects flame in video with negligible false-positive rates in real-time.",

keywords = "Deep Convolutional Generative Adversarial Neural Network, Fire detection, Flame detection",

author = "S{\"u}leyman Aslan and Uğur G{\"u}d{\"u}kbay and T{\"o}reyin, {B. Uğur} and {\c C}etin, {A. Enis}",

note = "Publisher Copyright: {\textcopyright} 2020, Springer Nature Switzerland AG.; 12th International Conference on Computational Collective Intelligence, ICCCI 2020 ; Conference date: 30-11-2020 Through 03-12-2020",

year = "2020",

doi = "10.1007/978-3-030-63007-2_63",

language = "English",

isbn = "9783030630065",

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

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "807--815",

editor = "Nguyen, {Ngoc Thanh} and Nguyen, {Ngoc Thanh} and Hoang, {Bao Hung} and Huynh, {Cong Phap} and Dosam Hwang and Bogdan Trawinski and Gottfried Vossen",

booktitle = "Computational Collective Intelligence - 12th International Conference, ICCCI 2020, Proceedings",

address = "Germany",

}

Aslan, S, Güdükbay, U, Töreyin, BU & Çetin, AE 2020, Deep Convolutional Generative Adversarial Networks for Flame Detection in Video. in NT Nguyen, NT Nguyen, BH Hoang, CP Huynh, D Hwang, B Trawinski & G Vossen (eds), Computational Collective Intelligence - 12th International Conference, ICCCI 2020, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12496 LNAI, Springer Science and Business Media Deutschland GmbH, pp. 807-815, 12th International Conference on Computational Collective Intelligence, ICCCI 2020, Da Nang, Viet Nam, 30/11/20. https://doi.org/10.1007/978-3-030-63007-2_63

Deep Convolutional Generative Adversarial Networks for Flame Detection in Video. / Aslan, Süleyman; Güdükbay, Uğur; Töreyin, B. Uğur et al.
Computational Collective Intelligence - 12th International Conference, ICCCI 2020, Proceedings. ed. / Ngoc Thanh Nguyen; Ngoc Thanh Nguyen; Bao Hung Hoang; Cong Phap Huynh; Dosam Hwang; Bogdan Trawinski; Gottfried Vossen. Springer Science and Business Media Deutschland GmbH, 2020. p. 807-815 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12496 LNAI).

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

TY - GEN

T1 - Deep Convolutional Generative Adversarial Networks for Flame Detection in Video

AU - Aslan, Süleyman

AU - Güdükbay, Uğur

AU - Töreyin, B. Uğur

AU - Çetin, A. Enis

PY - 2020

Y1 - 2020

N2 - Real-time flame detection is crucial in video-based surveillance systems. We propose a vision-based method to detect flames using Deep Convolutional Generative Adversarial Neural Networks (DCGANs). Many existing supervised learning approaches using convolutional neural networks do not take temporal information into account and require a substantial amount of labeled data. To have a robust representation of sequences with and without flame, we propose a two-stage training of a DCGAN exploiting spatio-temporal flame evolution. Our training framework includes the regular training of a DCGAN with real spatio-temporal images, namely, temporal slice images, and noise vectors, and training the discriminator separately using the temporal flame images without the generator. Experimental results show that the proposed method effectively detects flame in video with negligible false-positive rates in real-time.

AB - Real-time flame detection is crucial in video-based surveillance systems. We propose a vision-based method to detect flames using Deep Convolutional Generative Adversarial Neural Networks (DCGANs). Many existing supervised learning approaches using convolutional neural networks do not take temporal information into account and require a substantial amount of labeled data. To have a robust representation of sequences with and without flame, we propose a two-stage training of a DCGAN exploiting spatio-temporal flame evolution. Our training framework includes the regular training of a DCGAN with real spatio-temporal images, namely, temporal slice images, and noise vectors, and training the discriminator separately using the temporal flame images without the generator. Experimental results show that the proposed method effectively detects flame in video with negligible false-positive rates in real-time.

KW - Deep Convolutional Generative Adversarial Neural Network

KW - Fire detection

KW - Flame detection

UR - http://www.scopus.com/inward/record.url?scp=85097534330&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-63007-2_63

DO - 10.1007/978-3-030-63007-2_63

M3 - Conference contribution

AN - SCOPUS:85097534330

SN - 9783030630065

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

SP - 807

EP - 815

BT - Computational Collective Intelligence - 12th International Conference, ICCCI 2020, Proceedings

A2 - Nguyen, Ngoc Thanh

A2 - Hoang, Bao Hung

A2 - Huynh, Cong Phap

A2 - Hwang, Dosam

A2 - Trawinski, Bogdan

A2 - Vossen, Gottfried

PB - Springer Science and Business Media Deutschland GmbH

T2 - 12th International Conference on Computational Collective Intelligence, ICCCI 2020

Y2 - 30 November 2020 through 3 December 2020

ER -

Aslan S, Güdükbay U, Töreyin BU, Çetin AE. Deep Convolutional Generative Adversarial Networks for Flame Detection in Video. In Nguyen NT, Nguyen NT, Hoang BH, Huynh CP, Hwang D, Trawinski B, Vossen G, editors, Computational Collective Intelligence - 12th International Conference, ICCCI 2020, Proceedings. Springer Science and Business Media Deutschland GmbH. 2020. p. 807-815. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-63007-2_63

Deep Convolutional Generative Adversarial Networks for Flame Detection in Video

Abstract

Publication series

Conference

Bibliographical note

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this