Efficient and Compact Convolutional Neural Network Architectures for Non-temporal Real-time Fire Detection

Thomson, W.; Bhowmik, N.; Breckon, T.P.

Abstract

Automatic visual fire detection is used to complement traditional fire detection sensor systems (smoke/heat). In this work, we investigate different Convolutional Neural Network (CNN) architectures and their variants for the non-temporal real-time bounds detection of fire pixel regions in video (or still) imagery. Two reduced complexity compact CNN architectures (NasNet-A-OnFire and ShuffleNetV2-OnFire) are proposed through experimental analysis to optimise the computational efficiency for this task. The results improve upon the current state-of-the-art solution for fire detection, achieving an accuracy of 95% for full-frame binary classification and 97% for superpixel localisation. We notably achieve a classification speed up by a factor of 2.3× for binary classification and 1.3× for superpixel localisation, with runtime of 40 fps and 18 fps respectively, outperforming prior work in the field presenting an efficient, robust and real-time solution for fire region detection. Subsequent implementation on low-powered devices (Nvidia Xavier-NX, achieving 49 fps for full-frame classification via ShuffleNetV2-OnFire) demonstrates our architectures are suitable for various real-world deployment applications.

Citation

Thomson, W., Bhowmik, N., & Breckon, T. (2020, December). Efficient and Compact Convolutional Neural Network Architectures for Non-temporal Real-time Fire Detection. Presented at 19th IEEE International Conference on Machine Learning and Applications (ICMLA 2020), Miami, FL

• Presentation Conference Type: Conference Paper (published)
• Conference Name: 19th IEEE International Conference on Machine Learning and Applications (ICMLA 2020)
• Start Date: Dec 14, 2020
• End Date: Dec 17, 2020
• Acceptance Date: Sep 16, 2020
• Online Publication Date: Feb 23, 2021
• Publication Date: 2021
• Deposit Date: Oct 26, 2020
• Publicly Available Date: Oct 27, 2020
• Publisher: Institute of Electrical and Electronics Engineers
• DOI: https://doi.org/10.1109/icmla51294.2020.00030
• Public URL: https://durham-repository.worktribe.com/output/1141470