MS-YOLO: A Lightweight and High-Precision YOLO Model for Drowning Detection
Abstract
:1. Introduction
- To capture the subtle movements and posture changes in various aquatic environments while enhancing the identification of drowning behavior, a self-designed dual-residual MD-C2F module is put forward. Through the combination with dynamic convolution (DcConv), model complexity is lowered. By integrating multi-scale attention (EMA), the MDE-C2F module improves the outcome of small object detection in complex environments.
- A novel spatial pyramid structure (MSI-SPPF) based on SPPF is developed to enable the model to better understand the relationship between various aquatic backgrounds and drowning targets while reducing the false and missed detection caused by environmental factors.
- To remove the irrelevant drowning features while emphasizing the critical characteristics, the feature fusion algorithm is improved using a bi-directional feature pyramid network (BiFPN).
2. Datasets
3. Methods
3.1. YOLOv8 Improved Model
3.2. A Novel C2F Architecture
3.2.1. MD-C2F
3.2.2. MDE-C2F
3.2.3. MSI-SPPF
3.2.4. BiFPN
4. Experiments and Discussion
4.1. Experimental Environment and Configuration
4.2. Evaluation Metrics
4.3. Experimental Results and Analysis
4.3.1. Analysis of Model Training Results
4.3.2. Ablation Experiments
4.3.3. Comparison of Detection Results on the Self-Made Dataset
5. Conclusions and Future Work
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Acknowledgments
Conflicts of Interest
References
- World Health Organization (WHO). Available online: https://www.who.int/publications-detail-redirect/9789240046726 (accessed on 23 October 2023).
- Lanagan-Leitzel, L.K.; Skow, E.; Moore, C.M. Great expectations: Perceptual challenges of visual surveillance in lifeguarding. Appl. Cogn. Psychol. 2015, 29, 425–435. [Google Scholar]
- Laxton, V.; Crundall, D. The effect of lifeguard experience upon the detection of drowning victims in a realistic dynamic visual search task. Appl. Cogn. Psychol. 2018, 32, 14–23. [Google Scholar] [CrossRef]
- Lei, F.; Zhu, H.; Tang, F.; Wang, X. Drowning behavior detection in swimming pool based on deep learning. Signal Image Video Process. 2022, 16, 1683–1690. [Google Scholar] [CrossRef]
- Salehi, N.; Keyvanara, M.; Monadjemmi, S.A. An automatic video-based drowning detection system for swimming pools using active contours. Int. J. Image Graph. Signal Process. 2016, 8, 1–8. [Google Scholar] [CrossRef]
- Jalalifar, S.; Kashizadeh, A.; Mahmood, I.; Belford, A.; Drake, N.; Razmjou, A.; Asadnia, M. A smart multi-sensor device to detect distress in swimmers. Sensors 2022, 22, 1059. [Google Scholar] [CrossRef]
- Misiurewicz, J.; Bruliński, K.; Klembowski, W.; Kulpa, K.S.; Pietrusiewicz, J. Multipath propagation of acoustic signal in a swimming pool—Source localization problem. Sensors 2022, 22, 1162. [Google Scholar] [CrossRef]
- Liu, T.; He, X.; He, L.; Yuan, F. A video drowning detection device based on underwater computer vision. IET Image Process. 2023, 17, 1905–1918. [Google Scholar] [CrossRef]
- Kharrat, M.; Wakuda, Y.; Koshizuka, N.; Sakamura, K. Near drowning pattern recognition using neural network and wearable pressure and inertial sensors attached at swimmer’s chest level. In Proceedings of the 2012 19th International Conference on Mechatronics and Machine Vision in Practice (M2VIP), Fukuoka, Japan, 28–30 November 2012; IEEE: New York, NY, USA, 2012. [Google Scholar]
- Claesson, A.; Schierbeck, S.; Hollenberg, J.; Forsberg, S.; Nordberg, P.; Ringh, M.; Olausson, M.; Jansson, A.; Nord, A. The use of drones and a machine-learning model for recognition of simulated drowning victims—A feasibility study. Resuscitation 2020, 156, 196–201. [Google Scholar] [CrossRef]
- Alotaibi, A. Automated and intelligent system for monitoring swimming pool safety based on the IoT and transfer learning. Electronics 2020, 9, 2082. [Google Scholar] [CrossRef]
- Xie, X.; Cheng, G.; Wang, J.; Yao, X.; Han, J. Oriented R-CNN for object detection. In Proceedings of the IEEE/CVF International Conference on Computer Vision, Fukuoka, Japan, 11–17 October 2021. [Google Scholar]
- Girshick, R.; Donahue, J.; Darrell, T.; Malik, J. Rich feature hierarchies for accurate object detection and semantic segmenta tion. In Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Columbus, OH, USA, 23–28 June 2014. [Google Scholar]
- Wang, J.; Bi, J.; Wang, L.; Wang, X. A non-reference evaluation method for edge detection of wear particles in ferrograph Bimages. Mech. Syst. Signal Process. 2018, 100, 863–876. [Google Scholar] [CrossRef]
- He, K.; Zhang, X.; Ren, S.; Sun, J. Spatial pyramid pooling in deep convolutional networks for visual recognition. IEEE Trans. Pattern Anal. Mach. Intell. 2015, 37, 1904–1916. [Google Scholar] [CrossRef] [PubMed]
- Girshick, R. Fast R-CNN. In Proceedings of the IEEE International Conference on Computer Vision (ICCV), Santiago, Chile, 7–13 December 2015; pp. 1440–1448. [Google Scholar]
- Li, X.; Shang, M.; Qin, H.; Chen, L. Fast accurate fish detection and recognition of underwater images with Fast R-CNN. In Proceedings of the OCEANS 2015—MTS/IEEE Conference, Washington, DC, USA, 19–22 October 2015; pp. 1–5. [Google Scholar]
- Ren, S.; He, K.; Girshick, R.; Sun, J. Faster R-CNN: Towards real-time object detection with region proposal networks. IEEE Trans. Pattern Anal. Mach. Intell. 2017, 39, 1137–1149. [Google Scholar] [CrossRef]
- Mao, R.; Zhang, Y.; Wang, Z. Recognizing stripe rust and yellow dwarf of wheat using improved Faster-RCNN. Trans. Chin. Soc. Agric. Eng. (Trans. CSAE) 2022, 38, 176–185. [Google Scholar]
- He, B.; Cui, C.; Guo, W.; Du, L.; Liu, B.; Tang, Y. Ferrography wear particle recognition of gearbox based on Faster R-CNN. Lubr. Eng. 2020, 45, 105–112. [Google Scholar]
- Li, H.; Yan, K.; Jing, H.; Hou, R.; Liang, X.-H. Apple leaf pathology detection and recognition based on improved SSD. Sens. Microsyst. 2022, 41, 134–137. [Google Scholar]
- Hu, K.; Luo, R.-M.; Liu, Z.-Q.; Cao, Y.-F.; Liao, F.; Wang, W.-X.; Li, Q.; Sun, D.-Z. Detection of bergamot diseases and pests based on improved SSD. J. Nanjing Agric. Univ. 2023, 46, 813–821. [Google Scholar]
- Liu, W.; Anguelov, D.; Erhan, D.; Szegedy, C.; Reed, S.; Fu, C.-Y.; Berg, A.C. SSD: Single Shot MultiBox Detector. In Lecture Notes in Computer Science; Springer International Publishing: Berlin/Heidelberg, Germany, 2016; pp. 21–37. ISBN 978-3-319-46448-0. [Google Scholar]
- Peng, H.; Xue, C.; Shao, Y.; Chen, K.; Liu, H.; Xiong, J.; Chen, H.; Gao, Z.; Yang, Z. Litchi detection in the field using an improved YOLOv3 model. Int. J. Agric. Biol. Eng. 2022, 15, 211–220. [Google Scholar] [CrossRef]
- Sun, D.-Z.; Liu, H.; Liu, J.-Y.; Ding, Z.; Xie, J.-X.; Wang, W.-X. Recognition of tea diseases based on improved YOLOv4 model. J. Northwest A F Univ. (Nat. Sci. Ed.) 2023, 51, 145–154. [Google Scholar]
- Wang, G.; Chen, Y.; An, P.; Hong, H.; Hu, J.; Huang, T. UAV-YOLOv8: A small-object-detection model based on improved YOLOv8 for UAV aerial photography scenarios. Sensors 2023, 23, 7190. [Google Scholar] [CrossRef]
- Zhang, C.; Zhang, G.; Li, H.; Liu, H.; Tan, J.; Xue, X. Underwater target detection algorithm based on improved YOLOv4 with SemiDSConv and FIoU loss function. Front. Mar. Sci. 2023, 10, 1153416. [Google Scholar] [CrossRef]
- Jensen, M.B.; Gade, R.; Moeslund, T.B. Swimming pool occupancy analysis using deep learning on low quality video. In Proceedings of the 1st International Workshop on Multimedia Content Analysis in Sports, Seoul, Republic of Korea, 26 October 2018. [Google Scholar]
- Niu, Q.; Wang, Y.; Yuan, S.; Li, K.; Wang, X. An indoor pool drowning risk detection method based on improved YOLOv4. In Proceedings of the 2022 IEEE 5th Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC), Chongqing, China, 8–10 July 2022; IEEE: New York, NY, USA, 2022. [Google Scholar]
- Yang, R.; Wang, K.; Yang, L. An improved YOLOv5 algorithm for drowning detection in the indoor swimming pool. Appl. Sci. 2023, 14, 200. [Google Scholar] [CrossRef]
- Carballo-Fazanes, A.; Bierens, J.J.; The International Expert Group to Study Drowning Behaviour. The visible behaviour of drowning persons: A pilot observational study using analytic software and a nominal group technique. Int. J. Environ. Res. Public Health 2020, 17, 6930. [Google Scholar] [CrossRef] [PubMed]
- Kerdvibulvech, C. Human Hand Motion Recognition Using an Extended Particle Filter. In AMDO 2014, LNCS 8563; Perales, F.J., Santos-Victor, J., Eds.; Springer International Publishing: Cham, Switzerland, 2014; Volume 8563, pp. 71–80. [Google Scholar]
- Guo, M.-H.; Xu, T.-X.; Liu, J.-J.; Liu, Z.-N.; Jiang, P.-T.; Mu, T.-J.; Zhang, S.-H.; Martin, R.R.; Cheng, M.-M.; Hu, S.-M. Attention mechanisms in computer vision: A survey. Comput. Vis. Media 2022, 8, 331–368. [Google Scholar] [CrossRef]
- Ouyang, D.; He, S.; Zhang, G.; Luo, M.; Guo, H.; Zhan, J.; Huang, Z. Efficient multi-scale attention module with cross-spatial learning. In Proceedings of the ICASSP 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Rhodes, Greece, 4–10 June 2023; IEEE: New York, NY, USA, 2023. [Google Scholar]
- Hao, W.; Ren, C.; Han, M.; Zhang, L.; Li, F.; Liu, Z. Cattle Body Detection Based on YOLOv5-EMA for Precision Livestock Farming. Animals 2023, 13, 3535. [Google Scholar] [CrossRef] [PubMed]
- Bochkovskiy, A.; Wang, C.-Y.; Liao, H.-Y.M. Yolov4: Optimal speed and accuracy of object detection. arXiv 2020, arXiv:2004.10934. [Google Scholar]
- Chen, L.-C.; Papandreou, G.; Kokkinos, I.; Murphy, K.; Yuille, A.L. Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs. IEEE Trans. Pattern Anal. Mach. Intell. 2017, 40, 834–848. [Google Scholar] [CrossRef]
- Wang, X.; Girshick, R.; Gupta, A.; He, K. Non-local neural networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Salt Lake City, UT, USA, 18–22 June 2018. [Google Scholar]
- Yan, J.; Zhou, Z.; Zhou, D.; Su, B.; Xuanyuan, Z.; Tang, J.; Lai, Y.; Chen, J.; Liang, W. Underwater object detection algorithm based on attention mechanism and cross-stage partial fast spatial pyramidal pooling. Front. Mar. Sci. 2022, 9, 1056300. [Google Scholar] [CrossRef]
- Ge, Z.; Wang, C.-Y.; Liao, H.-Y.M. YOLOX: Exceeding YOLO series in 2021. arXiv 2021, arXiv:2107.08430. [Google Scholar]
- Qu, Z.; Gao, L.-Y.; Wang, S.-Y.; Yin, H.-N.; Yi, T.-M. An improved YOLOv5 method for large objects detection with multi-scale feature cross-layer fusion network. Image Vis. Comput. 2022, 125, 104518. [Google Scholar] [CrossRef]
- Li, C.; Li, L.; Jiang, H.; Weng, K.; Geng, Y.; Li, L.; Ke, Z.; Li, Q.; Cheng, M.; Nie, W.; et al. YOLOv6: A Single-Stage Object Detection Framework for Industrial Applications. arXiv 2022, arXiv:2209.02976. [Google Scholar]
- Hao, W.; Zhang, L.; Liu, Z.; Wang, X. YOLOv10: Real-Time End-to-End Object Detection. arXiv 2024, arXiv:2405.14458. [Google Scholar]
- Wang, C.Y.; Yeh, I.H.; Liao, H.-Y.M. YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information. arXiv 2024, arXiv:2402.13616. [Google Scholar]
- Jocher, G. YOLOv8 by Ultralytics. 2023. Available online: https://github.com/ultralytics/ultralytics (accessed on 15 February 2023).
Configuration Item | Version | Parameters |
---|---|---|
Operating System | Linux | |
CPU | AMD EPYC 7542 | 42 cores |
GPU | NVIDIA GeForce RTX3090 | 24 GB |
Deep Learning Framework | Pytorch | 2.2.1 |
Programming Language | Python | 3.9 |
Computing Architecture | CUDA | 12.1 |
Training Parameters | Values |
---|---|
Image size | 640 × 640 |
Epochs | 300 |
Batch | 64 |
Works | 16 |
Learning rate | 0.01 |
Optimizer | SGD |
Cache | True |
Close music | 10 |
BiFPN | MSI-SPPF | MD-C2F and MDE-C2F | P (%) | R (%) | MAP@0.5 (%) | F1 (%) | GFLOPs (G) | ||
---|---|---|---|---|---|---|---|---|---|
Drowning | Safety | All | |||||||
75.7 | 71.7 | 77.2 | 69.7 | 73.4 | 73 | 8.1 | |||
√ | 76 | 72.8 | 78 | 70.3 | 74.2 | 74 | 8.1 | ||
√ | 78.5 | 73.9 | 82.4 | 73.8 | 78 | 76 | 8.7 | ||
√ | 78.8 | 75.4 | 83.2 | 75.1 | 79.8 | 77 | 7.0 | ||
√ | √ | 79.9 | 77.2 | 82.5 | 74.2 | 78.4 | 78 | 8.6 | |
√ | √ | √ | 80.9 | 77.8 | 86.4 | 77.7 | 82 | 79 | 7.3 |
Algorithm | Precision (%) | Recall (%) | mAP@0.5 (%) | mAP@0.5:0.95 (%) | F1 (%) | GFLOPs (G) | ||
---|---|---|---|---|---|---|---|---|
Drowning | Safety | All | ||||||
Faster-RCNN | 64.8 | 70.2 | 68.3 | 56.97 | 62.63 | 28.9 | 67 | 170 |
SSD | 84.61 | 61.24 | 71.06 | 51.56 | 61.31 | 28.4 | 71 | 31.4 |
YOLOv6n | 78 | 76.1 | 85.6 | 77.3 | 81.4 | 36.1 | 78 | 11.8 |
YOLOv8n | 75.7 | 71.7 | 77.2 | 69.7 | 73.4 | 31 | 73 | 8.1 |
YOLOv9n | 78.2 | 72 | 84.7 | 76.9 | 80.8 | 36.1 | 78 | 313.4 |
YOLOv10n | 77.3 | 75.6 | 82.5 | 76 | 79.2 | 34 | 76 | 8.2 |
MS-YOLO | 80.9 | 77.8 | 86.4 | 77.7 | 82 | 36.5 | 79 | 7.3 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Song, Q.; Yao, B.; Xue, Y.; Ji, S. MS-YOLO: A Lightweight and High-Precision YOLO Model for Drowning Detection. Sensors 2024, 24, 6955. https://doi.org/10.3390/s24216955
Song Q, Yao B, Xue Y, Ji S. MS-YOLO: A Lightweight and High-Precision YOLO Model for Drowning Detection. Sensors. 2024; 24(21):6955. https://doi.org/10.3390/s24216955
Chicago/Turabian StyleSong, Qi, Bodan Yao, Yunlong Xue, and Shude Ji. 2024. "MS-YOLO: A Lightweight and High-Precision YOLO Model for Drowning Detection" Sensors 24, no. 21: 6955. https://doi.org/10.3390/s24216955
APA StyleSong, Q., Yao, B., Xue, Y., & Ji, S. (2024). MS-YOLO: A Lightweight and High-Precision YOLO Model for Drowning Detection. Sensors, 24(21), 6955. https://doi.org/10.3390/s24216955