A modified YOLOv5 helmet detection algorithm based on Swin Transformer
DOI:
https://doi.org/10.54097/fcis.v3i2.6914Keywords:
Safety helmet detection, YOLOv5, Swin TransformerAbstract
For the current stage of helmet detection in complex environments with low accuracy, missed detection and not easy to manage wearing, this paper proposes a YOLOv5 face helmet detection algorithm based on Swin Transformer improvement from the overall semantics of the image. In this paper, experiments are conducted using a self-built dataset to further enhance the performance of the model and improve the accuracy of face helmet detection through Mosaic data enhancement, label smoothing processing, adaptive weighted features combined with Wconcat module and the application of C3TR and C3STR modules to fuse multi-scale information, enhance the feature extraction capability of the network, and improve the generalization and robustness of the model with a self-built dataset . Experiments show that the improved YOLOv5 face helmet detection algorithm mAP based on Swin Transformer improves 5.7% compared with Faster RCNN, 6.1% compared with YOLOV3, 5.3% compared with YOLOV4, and 1.6% compared with the original algorithm. It performs well in helmet face detection tasks in complex environments, achieving real-time detection and higher accuracy, while reducing missed detections.
Downloads
References
Shi Hui. Research on deep learning based construction site helmet wearing detection algorithm [D]. Wuhan Institute of Technology,2019.DOI:10.27381/d.cnki.gwlgu.2019.001638.
Yu Bo. Safety helmet detection based on intelligent video surveillance [D]. Hebei University of Technology, 2011.
Gkioxari G , Hariharan B , Girshick R , et al. R-CNNs for Pose Estimation and Action Detection[J]. Computer ence, 2014.J.-M. Chang, W.-T. Hsiao, J.-L. Chen, H.-C. Chao, Mobile relay stations navigation-based self-optimization handover mechanism in WiMAX Networks, in: Proc. 2009 International Conference on Ubiquitous Information Technologies & Applications, 2009.B. Smith, “An approach to graphs of linear forms (Unpublished work style),” unpublished.
Girshick R. Fast R-CNN[C]// International Conference on Computer Vision. IEEE Computer Society, 2015. J.G. Wilson, F.C. Fraser (Eds.), Handbook of Teratology, vols. 1-4, Plenum Press, New York, 1977-1978.
Ren S , He K , Girshick R , et al. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks[C]// NIPS. 2016. W. Strunk Jr., E.B. White, The Elements of Style, third ed., MacMillan, New York, 1979 (Chapter 4).
Redmon J , Farhadi A . YOLO9000: Better, Faster, Stronger[C]// IEEE Conference on Computer Vision & Pattern Recognition. IEEE, 2017:6517-6525. Cancer Research UK, Cancer statistics reports for the UK2003 (accessed 13.03.03).
http://www.cancerresearchuk.org/aboutcancer/statistics/cancerstatsreport/
Redmon J, Farhadi A.YOLOv3: An Incremental Improvement[J]. arXiv e-prints, 2018.M. Young, The Techincal Writers Handbook. Mill Valley, CA: University Science, 1989.
Bochkovskiy A , Wang C Y , Liao H . YOLOv4: Optimal Speed and Accuracy of Object Detection[J]. 2020.
Wang J , Chen Y , Gao M , et al. Improved YOLOv5 network for real-time multi-scale traffic sign detection[J]. 2021.
Ding W-L, Fei Sh-Min. Research on helmet detection method based on improved YOLOv3 [J]. Electronic Testing, 2022.
Jin Yufang, Wu Xiang, Dong Hui, et al. Improved helmet wearing detection algorithm based on YOLO v4 [J]. Computer Science, 2021.
Yue H, Huang H, Lin MH, Gao M, Li Y, Chen L. Safety helmet wearing detection based on improved YOLOv5 [J]. Computers and Modernization,2022(06):104-108+126.
Qin ZH, Lei M, Song WG, Zhang W. Safety helmet detection method based on lightweight deep learning model[J]. Science Technology and Engineering,2022,22(14):5659-5665.
Vaswani A , Shazeer N , Parmar N , et al. Attention Is All You Need[C]// arXiv. arXiv, 2017.
Liu Z, Lin Y, Cao Y, et al. Swin Transformer: Hierarchical Vision Transformer using Shifted Windows[J]. 2021.
Downloads
Published
Issue
Section
How to Cite
