Comparative Performance Evaluation of MobileNetV3 and ResNet50 for Forest Fire Image Classification
DOI:
10.33395/sinkron.v9i4.15415Keywords:
Deep Learning, MobileNetV3, ResNet50, Image Classification, Forest FiresAbstract
Indonesia is one of the countries with a high incidence of forest and land fires (karhutla), especially during the dry season, thus requiring a fast and efficient early detection system. This study aims to compare the performance of two popular deep learning architectures, namely MobileNetV3 (Large and Small variants) and ResNet50, in forest fire image classification tasks using a transfer learning-based approach. This study emphasizes the comparison between accuracy and computational efficiency in a CPU-only environment, which represents real-world conditions of use in the field without GPU support. The dataset used is a combination of local field images from the Puncak area, Bogor, and a curated public forest fire dataset to ensure the model's generalization ability to diverse geographical conditions. The results of the experiment show that ResNet50 provides the highest accuracy with a training accuracy value of 0.677 and a validation accuracy of 0.647, but requires longer training and inference times. Meanwhile, MobileNetV3-Large and MobileNetV3-Small showed better computational efficiency, with only slightly lower accuracy (0.635 and 0.61) and high training stability. These findings confirm that lightweight models such as MobileNetV3 strike an optimal balance between accuracy, speed, and resource consumption, making them an ideal solution for implementing edge computing-based early detection systems. Overall, this research contributes by providing an empirical comparative analysis that can serve as a reference for selecting deep learning architectures for efficient and adaptive forest fire detection systems that are constrained by hardware limitations.
Downloads
References
Barmpoutis, P., Papaioannou, P., Dimitropoulos, K., & Grammalidis, N. (2020). A review on early forest fire detection systems using optical remote sensing. Sensors, 20(22), 6442. doi:10.3390/s20226442
Frizzi, S., Kaabi, R., Bouchouicha, M., Ginoux, J.-M., Moreau, E., & Fnaiech, F. (2016). Convolutional neural network for video fire and smoke detection. In IECON 2016-42nd Annual Conference of the IEEE Industrial Electronics Society (pp. 877–882). doi:10.1109/IECON.2016.7793196
Guede-Fernández, F., Martins, L., de Almeida, R. V., Gamboa, H., & Vieira, P. (2019). A deep learning based object identification system for forest fire detection. Fire, 2(4), 52. doi:10.3390/fire2040052
He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 770–778). doi:10.1109/CVPR.2016.90
Hindarto, D., & Kamal, I. (2021). Penerapan Transfer Learning Menggunakan EfficientNet untuk Klasifikasi Jenis Sampah Plastik. Jurnal Media Informatika Budidarma, 5(3), 998–1005. doi:10.30865/mib.v5i3.3053
Hindarto, D., & Putra, A. K. (2022). Optimasi Model Deep Learning pada Perangkat Edge untuk Pemantauan Lingkungan secara Real-Time. Journal of Intelligent Systems and Computing (JISC), 4(1), 45–56.
Hindarto, D., & Rahmawati, E. (2023). Lightweight Deep Learning Model for Plant Disease Detection on Mobile Devices. International Journal of Advanced Computer Science and Applications (IJACSA), 14(5), 112–119. doi:10.14569/IJACSA.2023.0140512
Hindarto, D., & Sari, Y. (2021). Implementasi Convolutional Neural Network untuk Deteksi Dini Kebakaran Hutan Berbasis Citra Drone. Jurnal SinkrOn, 6(2), 120–130. doi:10.33395/sinkron.v6i2.10987
Howard, A. G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T., ... & Adam, H. (2017). Mobilenets: Efficient convolutional neural networks for mobile vision applications. arXiv preprint arXiv:1704.04861.
Howard, A., Sandler, M., Chu, G., Chen, L.-C., Chen, B., Tan, M., ... & Vasudevan, V. (2019). Searching for mobilenetv3. In Proceedings of the IEEE/CVF International Conference on Computer Vision (pp. 1314–1324). doi:10.1109/ICCV.2019.00140
Hu, J., Shen, L., & Sun, G. (2018). Squeeze-and-excitation networks. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 7132–7141). doi:10.1109/CVPR.2018.00745
Khan, A., Hassan, B., Khan, S., & Ahmed, R. (2021). A comprehensive study of mobileNetV3 for fire detection in images. Journal of Real-Time Image Processing, 18, 2165–2177. doi:10.1007/s11554-021-01121-y
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). Imagenet classification with deep convolutional neural networks. Advances in neural information processing systems, 25.
Li, P., & Zhao, W. (2020). Image fire detection based on lightweight convolutional neural network. Journal of Physics: Conference Series, 1684(1), 012094. doi:10.1088/1742-6596/1684/1/012094
Muhammad, K., Ahmad, J., & Baik, S. W. (2018). Early fire detection using convolutional neural networks during surveillance for effective disaster management. Neurocomputing, 288, 30–42. doi:10.1016/j.neucom.2017.04.083
Park, M., & Tran, D. Q. (2021). Wildfire smoke detection based on deep learning approaches. Remote Sensing, 13(16), 3115. doi:10.3390/rs13163115
Sudhakar, S., & Sivakumar, V. (2021). A comparative analysis of ResNet and MobileNet for fire detection in surveillance videos. International Journal of Advanced Computer Science and Applications (IJACSA), 12(8), 1–8. doi:10.14569/IJACSA.2021.0120801
Yuan, F., Zhang, L., Xia, X., & Li, Y. (2021). Forest fire detection using deep learning and UAV imagery. IEEE Geoscience and Remote Sensing Letters, 19, 1–5. doi:10.1109/LGRS.2021.3056389
Zhao, Y., Ma, J., Li, X., & Zhang, J. (2021). Saliency detection for forest fire smoke using deep convolutional neural networks. Fire Technology, 57(2), 925–945. doi:10.1007/s10694-020-01014-9
Zheng, Z., & Wang, P. (2022). A lightweight network for forest fire smoke recognition combining SE module and MobileNetV3. Ecological Informatics, 69, 101641. doi:10.1016/j.ecoinf.2022.101641
El-Madafri I, Peña M, Olmedo-Torre N. The Wildfire Dataset: Enhancing Deep Learning-Based Forest Fire Detection with a Diverse Evolving Open-Source Dataset Focused on Data Representativeness and a Novel Multi-Task Learning Approach. Forests. 2023; 14(9):1697. https://doi.org/10.3390/f14091697
Downloads
|
How to Cite
Issue
Section
License
Copyright (c) 2025 Muhammad Rizky Amirullah Hidayat, Djarot Hindarto, Asrul Sani
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Google Scholar 
Moraref
PKP Index
Garuda Kemdikbud
Indonesia OneSearch
OCLC Worldcat
Dimensions
Index Copernicus
Scilit
Muhammad Khoiruddin harahap
