Abstract
To prevent the spread of the epidemic, a lightweight mask-wearing detection algorithm based on the modified YOLOv5 is proposed. The three categories of mask-wearing, mask-less, and non-standard mask-wearing were studied, and images of faces covered with hands and clothing were added as interference items to make the study more realistic. Using YOLOv5 as the base model, the modified deformable convolution is introduced into the CSP-Darknet53 module of the backbone network, which effectively enhances the model's ability to extract key information such as faces and masks and improves the detection accuracy of the model. In addition, image defogging based on dark channel prior is introduced to make the model adapt to detect whether pedestrians wear a mask correctly in special weather. Finally, the knowledge distillation method is used to perform light-weight compression on the model to solve the problem of slow detection speed in practical applications. Training and testing on the self-made T-Mask dataset, the mAP of the model reaches 93.6%, which is about 3% points higher than the original YOLOv5m algorithm, and the number of parameters is reduced to 1/3 of the original one, which has a higher practical value.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Ni, X., Xing, Y., Suo, J., et al.: Characteristics and functions of microbial aerosols in medical institutions. Chin. J. Hosp. Infect. 30(8), 1183–1190 (2020)
Wang, R., Zeng, Q., Liu, H.: Meta-analysis of the relationship between medical staff wearing masks and infections of respiratory infectious diseases. J. Environ. Health 29(3), 269–270 (2012)
Davies, A., Thompson, K.-A., Giri, K., et al.: Testing the efficacy of homemade masks: would they protect in an influenza pandemic? Disaster Med. Public Health Prep. 7(4), 413–418 (2013)
Macintyre, C.R., Cauchemez, S., Dwyer, D.E., et al.: Face mask use and control of respiratory virus transmission in households. Emerg. Infect. Dis. 15(2), 233 (2009)
Kux, L.: Department of health and human services. Food and Drug Administration [docket no FDA-2012-D-0146]: guidance for industry on irritable bowel syndrome-clinical evaluation of drugs for treatment: availability Fed Regist, vol. 77, pp. 32124–32125 (2012)
Henry, S., Manuguerra, J.-C., Escourolle, D., et al.: National plan for the prevention and control “influenza pandemic”. La Revue du Praticien 58(15), 1687–1693 (2008)
Suzhen, C., Dongsen, W., Xing, C., et al.: Research on Chinese residents wearing masks for protective behavior during the COVID-19 epidemic. Environ. Sci. Res. 33(7), 1649–1658 (2020)
Zuodong, N., Tao, Q., Handong, L., et al.: Improved RetinaFace mask wearing detection algorithm in natural scenes. Comput. Eng. Appl. 56(12), 1–7 (2020)
Wu, P., Li, H., Zeng, N., et al.: FMD-Yolo: an efficient face mask detection method for COVID-19 prevention and control in public. Image Vis. Comput. 117, 104341 (2022)
Kumar, A., Kalia, A., Sharma, A., et al.: A hybrid tiny YOLO v4-SPP module based improved face mask detection vision system. J. Ambient Intell. Humaniz. Comput., 1–14 (2021)
Xin, Z., Rui, L., Weiqi, L.: Mask wearing detection algorithm based on improved YOLOV3. Comput. Eng. Des. 43(05), 1319–1326 (2022)
Redmon, J., Divvala, S., Girshick, R., et al.: You only look once: Unified, real-time object detection. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (2016)
Redmon, J., Farhadi, A.: YOLO9000: better, faster, stronger. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (2017)
Redmon, J., Farhadi, A.: Yolov3: an incremental improvement. arXiv preprint arXiv:180402767 (2018)
Wang, C.-Y., Liao, H.-Y.M., Wu, Y.-H., et al.: CSPNet: a new backbone that can enhance learning capability of CNN. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (2020)
Dai, J., Qi, H., Xiong, Y., et al.: Deformable convolutional networks. In: Proceedings of the IEEE International Conference on Computer Vision (2017)
Zhu, X., Hu, H., Lin, S., et al.: Deformable ConvNets V2: more deformable, better results. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2019)
Gou, J., Yu, B., Maybank, S.J., et al.: Knowledge distillation: a survey. Int. J. Comput. Vis. 129(6), 1789–1819 (2021)
He, K., Sun, J., Tang, X.: Single image haze removal using dark channel prior. IEEE Trans. Pattern Anal. Mach. Intell. 33(12), 2341–2353 (2010)
Cabani, A., Hammoudi, K., Benhabiles, H., Melkemi, M.: MaskedFace-Net - a dataset of correctly/incorrectly masked face images in the context of COVID-19. Smart Health (2020). ISSN 2352-6483. https://doi.org/10.1016/j.smhl.2020.100144
Acknowledgement
This work is supported by Heilongjiang Provincial Natural Science Foundation of China (No. LH2022F035) and 2022 Harbin University of Commerce's the support program project of faculty “Innovation” project (No. XL0068) and Harbin University of Commerce Graduate Innovative Research Project (No. YJSCX2022-743HSD).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Su, X., Xu, H., Zhang, Y., Zhao, J., Zhang, F., Chen, X. (2023). Lightweight Mask Wearing Detection Algorithm Based on Improved YOLOv5. In: Hassanien, A., Rizk, R.Y., Pamucar, D., Darwish, A., Chang, KC. (eds) Proceedings of the 9th International Conference on Advanced Intelligent Systems and Informatics 2023. AISI 2023. Lecture Notes on Data Engineering and Communications Technologies, vol 184. Springer, Cham. https://doi.org/10.1007/978-3-031-43247-7_44
Download citation
DOI: https://doi.org/10.1007/978-3-031-43247-7_44
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-43246-0
Online ISBN: 978-3-031-43247-7
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)