CSAF-YOLO: An Improved Smoke and Fire Object Detection Algorithm for Complex Scenes Based on YOLOv11n
Abstract
Early detection of flames and smoke is crucial for fire warning systems, yet blurred smoke boundaries, small-scale targets, and complex background interference still limit existing methods. This paper proposes CSAF-YOLO, an improved detection model based on YOLOv11n. A C3-MCG module is introduced into the backbone to enhance multi-scale contextual feature representation. A DB-SAFM module is designed in the feature fusion stage to improve semantic and spatial feature alignment through a dual-branch self-attention mechanism. AWHIoU is adopted as the loss function, combined with a bi-phase focusing strategy and a scale-adaptive weighting mechanism to achieve adaptive optimisation for sampl0065s with different IoU qualities and targets of different scales. Experimental results show that CSAF-YOLO significantly improves detection performance over mainstream models, achieving 2.2% and 2.3% gains in precision and mAP50 over YOLOv11n, respectively, while maintaining a high FPS of 177.75, demonstrating strong potential for complex monitoring and real-time detection.References
[1] Los Angeles County (2025). Eaton & Palisades Fires After-Action Reviews. Available from: https://lacounty.gov/aar/ [Accessed 2026-02-05].
[2] United Nations Chile (2026). Incendios forestales 2026: Reporte de Situación N°2. Available from: https://chile.un.org/es/308852-incendios-forestales-2026-reporte-de-situaci%C3%B3n-n%C2%B02 [Accessed 2026-02-05].
[3] Özel, B., M.S. Alam,M.U. Khan (2024). Review of modern forest fire detection techniques: Innovations in image processing and deep learning. Information, 15(9), pp. 538. https://doi.org/10.3390/info15090538
[4] Ramadan, M.N.A., T. Basmaji, A. Gad, H. Hamdan, B.T. Akgün, M.A.H. Ali, M. Alkhedher,M. Ghazal (2024). Towards early forest fire detection and prevention using AI-powered drones and the IoT. Internet of Things, 27, pp. 101248. https://doi.org/10.1016/j.iot.2024.101248
[5] Gragnaniello, D., A. Greco, C. Sansone,B. Vento (2024). Fire and smoke detection from videos: A literature review under a novel taxonomy. Expert Systems with Applications, 255, pp. 124783. https://doi.org/10.1016/j.eswa.2024.124783
[6] Wang, X., M. Li, M. Gao, Q. Liu, Z. Li,L. Kou (2023). Early smoke and flame detection based on transformer. Journal of Safety Science and Resilience, 4(3), pp. 294-304. https://doi.org/10.1016/j.jnlssr.2023.06.002
[7] Fernandes, A.M., A.B. Utkin,P. Chaves (2023). Automatic early detection of wildfire smoke with visible-light cameras and EfficientDet. Journal of Fire Sciences, 41(4), pp. 122-135. https://doi.org/10.1177/07349041231163451
[8] Chaturvedi, S., C. Shubham Arun, P. Singh Thakur, P. Khanna,A. Ojha (2024). Ultra-lightweight convolution-transformer network for early fire smoke detection. Fire Ecology, 20(1), pp. 83. https://doi.org/10.1186/s42408-024-00304-9
[9] Kim, S.-Y.,A. Muminov (2023). Forest fire smoke detection based on deep learning approaches and unmanned aerial vehicle images. Sensors, 23(12), pp. 5702. https://doi.org/10.3390/s23125702
[10] Wang, H., X. Fu, Z. Yu,Z. Zeng (2025). DSS-YOLO: an improved lightweight real-time fire detection model based on YOLOv8. Scientific Reports, 15(1), pp. 8963. https://doi.org/10.1038/s41598-025-93278-w
[11] Ren, S., K. He, R. Girshick,J. Sun (2017). Faster R-CNN: Towards real-time object detection with region proposal networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(6), pp. 1137-1149. https://doi.org/10.1109/TPAMI.2016.2577031
[12] Liu, W., D. Anguelov, D. Erhan, C. Szegedy, S. Reed, C.-Y. Fu,A.C. Berg (2016). SSD: Single shot multibox detector. European conference on computer vision, Springer, pp. 21-37. https://doi.org/10.1007/978-3-319-46448-0_2
[13] Lin, T.Y., P. Goyal, R. Girshick, K. He,P. Dollár (2020). Focal Loss for Dense Object Detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, 42(2), pp. 318-327. https://doi.org/10.1109/TPAMI.2018.2858826
[14] Gonçalves, L.A.O., R. Ghali,M.A.J.F. Akhloufi (2024). YOLO-Based models for smoke and Wildfire Detection in Ground and aerial images. Fire, 7(4), pp. 140. https://doi.org/10.3390/fire7040140
[15] Khanam, R.,M. Hussain (2024). Yolov11: An overview of the key architectural enhancements. arXiv preprint arXiv:2410.17725, pp.
[16] Lin, T.-Y., P. Dollár, R. Girshick, K. He, B. Hariharan,S. Belongie (2017). Feature pyramid networks for object detection. Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 2117-2125. https://doi.org/10.1109/CVPR.2017.106
[17] Liu, S., L. Qi, H. Qin, J. Shi,J. Jia (2018). Path aggregation network for instance segmentation. Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 8759-8768. https://doi.org/10.1109/CVPR.2018.00913
[18] Tan, M., R. Pang,Q.V. Le (2020). EfficientDet: Scalable and efficient object detection. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp. 10781-10790. https://doi.org/10.1109/CVPR42600.2020.01079
[19] Hu, H., J. Gu, Z. Zhang, J. Dai,Y. Wei (2018). Relation Networks for Object Detection. Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 3588-3597. https://doi.org/10.1109/CVPR.2018.00378
[20] Wu, T., S. Tang, R. Zhang, J. Cao,Y. Zhang (2021). CGNet: A Light-Weight Context Guided Network for Semantic Segmentation. IEEE Transactions on Image Processing, 30, pp. 1169-1179. https://doi.org/10.1109/TIP.2020.3042065
[21] Hu, J., L. Shen,G. Sun (2018). Squeeze-and-Excitation Networks. Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 7132-7141. https://doi.org/10.1109/CVPR.2018.00745
[22] Woo, S., J. Park, J.-Y. Lee,I.S. Kweon (2018). CBAM: Convolutional Block Attention Module. Proceedings of the European conference on computer vision (ECCV), pp. 3-19. https://doi.org/10.1007/978-3-030-01234-2_1
[23] Ejaz, U., M.A. Hamza,H.-c. Kim (2025). Channel Attention for Fire and Smoke Detection: Impact of Augmentation, Color Spaces, and Adversarial Attacks. Sensors, 25(4), pp. 1140. https://doi.org/10.3390/s25041140
[24] Jin, C., A. Zheng, Z. Wu,C. Tong (2023). Real-time fire smoke detection method combining a self-attention mechanism and radial multi-scale feature connection. Sensors, 23(6), pp. 3358. https://doi.org/10.3390/s23063358
[25] Tolstikhin, I.O., N. Houlsby, A. Kolesnikov, L. Beyer, X. Zhai, T. Unterthiner, J. Yung, A. Steiner, D. Keysers,J. Uszkoreit (2021). MLP-Mixer: An all-mlp architecture for vision. Advances in neural information processing systems, 34, pp. 24261-24272.
[26] Rezatofighi, H., N. Tsoi, J. Gwak, A. Sadeghian, I. Reid,S. Savarese (2019). Generalized intersection over union: A metric and a loss for bounding box regression. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp. 658-666. https://doi.org/10.1109/CVPR.2019.00075
[27] Zheng, Z., P. Wang, W. Liu, J. Li, R. Ye,D. Ren (2020). Distance-IoU loss: Faster and better learning for bounding box regression. Proceedings of the AAAI conference on artificial intelligence, New York, NY, USA, pp. 12993-13000. https://doi.org/10.1609/aaai.v34i07.6999
[28] Song, B., S. Zhao, Z. Wang, J. Chen, W. Liu,X. Liu (2025). LW-DETR: a lightweight transformer-based object detection algorithm for efficient railway crossing surveillance. Complex & Intelligent Systems, 11(12), pp. 480. https://doi.org/10.1007/s40747-025-02111-4
[29] Xu, L., Y. Zhao, Y. Zhai, L. Huang,C. Ruan (2024). Small Object Detection in UAV Images Based on YOLOv8n. International Journal of Computational Intelligence Systems, 17(1), pp. 223. https://doi.org/10.1007/s44196-024-00632-3
[30] de Venâncio, P.V.A.B., A.C. Lisboa,A.V. Barbosa (2022). An automatic fire detection system based on deep convolutional neural networks for low-power, resource-constrained devices. Neural Computing and Applications, 34(18), pp. 15349-15368. https://doi.org/10.1007/s00521-022-07467-z
DOI:
https://doi.org/10.31449/inf.v50i14.14573Keywords:
Fire smoke detection, YOLOv11n, Multi-scale contextual modelling, Bi-branch self-attention feature fusionDownloads
Published
Issue
Section
License
Authors retain copyright in their work. By submitting to and publishing with Informatica, authors grant the publisher (Slovene Society Informatika) the non-exclusive right to publish, reproduce, and distribute the article and to identify itself as the original publisher.
All articles are published under the Creative Commons Attribution license CC BY 3.0. Under this license, others may share and adapt the work for any purpose, provided appropriate credit is given and changes (if any) are indicated.
Authors may deposit and share the submitted version, accepted manuscript, and published version, provided the original publication in Informatica is properly cited.
