• Sorted by Date • Classified by Publication Type • Classified by Research Category •
Süleyman Aslan, Uğur Güdükbay, B. Uğur Töreyin, and A. Enis Çetin. Deep Convolutional Generative Adversarial Networks for Flame Detection in Video. In Computational Collective Intelligence, pp. 807–815, CCI '20, Lecture Notes in Computer Science (LNCS) 12496, Springer International Publishing, Cham, November 2020.
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.
@InProceedings{AslanEtAl2020, author="S{\"u}leyman Aslan and U{\^g}ur G{\"u}d{\"u}kbay and B. U{\^g}ur T{\"o}reyin and A. Enis {\c C}etin", editor="Nguyen, Ngoc Thanh and Hoang, Bao Hung and Huynh, Cong Phap and Hwang, Dosam and Trawi{\'{n}}ski, Bogdan and Vossen, Gottfried", title="Deep Convolutional Generative Adversarial Networks for Flame Detection in Video", booktitle="Computational Collective Intelligence", series = {CCI '20, Lecture Notes in Computer Science (LNCS)}, volume = 12496, year="2020", month = {November}, publisher="Springer International Publishing", address="Cham", pages="807--815", 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.", isbn="978-3-030-63007-2", bib2html_dl_pdf = "http://www.cs.bilkent.edu.tr/~gudukbay/publications/papers/conf_papers/Aslan_Et_Al_ICCCI_2020.pdf", bib2html_pubtype = {Refereed Conference Papers}, bib2html_rescat = {Computer Vision} }
Generated by bib2html.pl (written by Patrick Riley ) on Sun Apr 21, 2024 11:32:41