Сравнение эффективности алгоритмов YOLOv5 и YOLOv8 для обнаружения средств индивидуальной защиты человека

S A Filichkin; S V Vologdin

doi:10.22213/2410-9304-2023-3-124-131

Authors

S. A. Filichkin Kalashnikov Izhevsk State Technical University
S. V. Vologdin Kalashnikov Izhevsk State Technical University

DOI:

https://doi.org/10.22213/2410-9304-2023-3-124-131

Keywords:

yolov8, YOLOv5, neural network, pattern recognition, computer vision

Abstract

The article deals with the comparative analysis of different architectures of convolutional neural networks of YOLO class: YOLOv5 and YOLOv8. Neural network was trained on a dataset consisting of more than 2,200 marked digital images. Detection and recognition of 8 classes of personal protective equipment such as medical masks, gloves, helmets, goggles, uniforms and others was performed. To evaluate the efficiency of the object detection algorithms such characteristics as accuracy metrics, learning time and memory size of the neural networks were used. The research confirms the high efficiency and accuracy of the YOLO algorithms in the detection and recognition of objects in digital still and video images. The research shows that the average accuracy of mAP object detection in YOLOv8 is 3% higher than in YOLOv5, while the performance in the new algorithm has increased by more than 50% compared with the earlier version of the neural network. The results obtained can be used to improve the object detection system in various fields, such as the automotive industry, medical and scientific research, the security field, and etc. Based on the results of the experiment, conclusions were made regarding the selection of an algorithm for the detection of human personal protective equipment. Further research in this area may be based on expanding the volume of training datasets to improve the accuracy of object recognition and evaluate the performance of algorithms on large amounts of data. Further research is planned in the area of optimizing the architecture of convolutional neural networks to improve the efficiency, speed and accuracy of object detection in digital images.

Author Biographies

S. A. Filichkin, Kalashnikov Izhevsk State Technical University

Post-graduate

S. V. Vologdin, Kalashnikov Izhevsk State Technical University

DSc. in Engineering, Associate Professor

References

Рябов П. И., Вологдин С. В., Максимова В. В. / Алгоритм распознавания изображений с приборов учета электроэнергии // Интеллектуальные системы в производстве. 2017. Т. 15, № 4. С. 42-48.

Vologdin S.V., Shushkov I.D. and Bysygin E.K. Portable data collection terminal in the automated power consumption measurement system. Journal of Physics: Conference Seriesthis link is disabled, 2018. DOI 10.1088/1742-6596/944/1/012122.

Sharipov, R., Panchenko, O. Problems of Developing User Identification Systems by Keyboard Handwriting. In: Radionov, A.A., Gasiyarov, V.R. (eds) Advances in Automation IV.RusAutoCon 2022. Lecture Notes in Electrical Engineering, vol 986. Springer, 2023.

Ziyang Tang, Xiang Liu, Guangyu Shen, Baijian Yang. Penet: object detection using points estimation in aerial im ages. arXiv preprint arXiv:2001.08247, 2020.

Roman Solovyev, Weimin Wang, Tatiana Gabruseva. Weighted boxes fusion: Ensembling boxes from different object detection models. Image and Vision Computing, 107:104117, 2021.

Sanghyun Woo, Jongchan Park, Joon-Young Lee, In So Kweon. Cbam: Convolutional block attention module. In Proceedings of the European conference on computer vision (ECCV), 2018, pp. 3-19.

Shu Liu, Lu Qi, Haifang Qin, Jianping Shi, and Jiaya Jia. Path aggregation network for instance segmentation. In Pro ceedings of the IEEE conference on computer vision and pat tern recognition, 2018, pp. 8759-8768.

Qiang Chen, Yingming Wang, Tong Yang, Xiangyu Zhang, Jian Cheng, Jian Sun. You Only Look One-level Feature. arXiv:2103.09460, 2021.

Zhi Zhang, Tong He, Hang Zhang, Zhongyue Zhang, Junyuan Xie, Mu Li. Bag of Freebies for Training Object Detection Neural Networks. arXiv:1902.04103, 2019.

Xingkui Zhu, Shuchang Lyu, Xu Wang, Qi Zhao. Improved YOLOv5 Based on Transformer Prediction Head for Object Detection on Drone-captured Scenarios. arXiv:2108.11539, 2021.

Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Syl vain Gelly, Jakob Uszkoreit, and Neil Houlsby. An image is worth 16x16 words: Transformers for image recognition at scale. In 9th International Conference on Learning Rep resentations, ICLR 2021, 2021.

Ultralytics YOLOv8. URL: https://github.com/ultralytics/ultralytics (дата обращения 20.02.2023).

Tsung-Yi Lin, Piotr Dollar, Ross Girshick, Kaiming He, Bharath Hariharan, and Serge Belongie. Feature pyra mid networks for object detection. In Proceedings of the IEEE conference on computer vision and pattern recogni tion, 2017, pp. 2117-2125.

Yang, G., Feng, W., Jin, J., Lei, Q., Li, X., Gui, G., Wang, W. Face mask recognition system with YOLOV5 based on image recognition. In 2020 IEEE 6th International Conference on Computer and Communications (ICCC), 2020, pp. 1398-1404.

Ieamsaard, J., Charoensook, S. N., Yammen, S. Deep learning-based face mask detection using yolov5. In 2021 9th International Electrical Engineering Congress (iEECON), 2021, pp. 428-431.

Применение нейронной сети YOLOv5 для контроля соблюдения регламентов персонала на предприятии / С. А. Филичкин, С. В Вологдин // Информационные технологии в науке, промышленности и образовании: сборник трудов Всероссийской научно-технической конференции. 2022. С. 73-80.

Girshick, R., Donahue, J., Darrell, T., Malik, J. Rich feature hierarchies for accurate object detection and semantic segmentation. In Proceedings of the IEEE conference on computer vision and pattern recognition, 2014, pp. 580-587.

Du, Y., Pan, N., Xu, Z., Deng, F., Shen, Y., Kang, H. Pavement distress detection and classification based on YOLO network.International Journal of Pavement Engineering, 2021, pp. 1659-1672. DOI 10.1080/10298436.2020.1714047.

Alexey Bochkovskiy, Chien-Yao Wang, HongYuan Mark Liao. Yolov4: Optimal speed and accuracy of object detection. arXiv preprint arXiv:2004.10934, 2020.

Vorozhtsova, N., Shushkov, I., Vologdin, S. System approach to development of intellectual information mobile system for electric power metering. E3S Web of Conferences, 2019, pp. 114. DOI 10.1088/1757-899X/1047/1/012096.