Real-time Pedestrian Tracking Based on YOLOv3 and Prototype Clustering


  • Ruopeng Li



Real-time Pedestrian Tracking, YOLO Network, Lightweight model.


Although the accuracy of existing neural network models is high, in pedestrian tracking tasks, due to the uncertainty of targets, when tracking new targets, it is necessary to fine-tune the model, which further requires large computing and storage resource overhead. Therefore, its application on some lightweight platforms, such as robots and UAVs, is limited. Pedestrian tracking by robots and UAVs still faces great challenges in occlusion, multi-target, target loss, etc. This paper mainly solves the problem of real-time pedestrian tracking by object detection model of robot lightweight, which is mainly based on YOLO network to detect pedestrians, and then proposes a novel lightweight model and prototype clustering algorithm. Numerous experiments on the ETH dataset validate the superiority and effectiveness of our approach.


Download data is not yet available.


An Feng-Ping, Liu Jun-E, Pedestrian Re-identification Algorithm Based on Visual Attention-positive Sample Generation Network Deep Learning Model..Information Fusion. Volume 86-87, Issue . 2022. PP 136-145

Zheng Feilong, Research on Person Re-ID Algorithm Based on Dynamic Images, Information and Computer (Theoretical Edition), 2019(03)

Chen Yang. Research on Video-based Person Re-ID Method. Huazhong University of Science and Technology, 2019

Luo Hao, Jiang Wei, Fan Xing et al. Research Progress of Person Re-ID Based on Deep Learning. Chinese Journal of Automation, 2019, 45(11): 2032-2049.

Huang Haixin, Tao Wenbo, Du Tingting, Overview of Person Re-ID Based on Metric Learning, Journal of Shenyang University of Technology, 2023, 42(05)

Tang Jiamin, Han Hua, Huang Li, Coarse and Fine Grain Feature Extraction Based on Unsupervised Learning in Pedestrian Recognition, Computer Engineering. 2022, 48(04)

Chang Zhe, Research on Pedestrian Re-recognition Algorithm for Complex Scenarios, Central South University, 2022

Zhao Cairong, Qi Ding, Key Technologies of Intelligent Video Surveillance: A Review of Pedestrian Re-recognition Research. Science in China: Information Science, 2021, 51(12).

Geng Yanbing, Lian Yongjian, Cross-resolution Person Re-ID Based on Multi-granularity Feature Generation Adversarial Networks, Computer Applications, 2022, 42(11)

Chen Peixian, Research on Domain Migration Methods for Person Re-ID, Xiamen University, 2021.

Bian Ziyang, Li Jianan, Cross-modal Person Re-ID Method and Application Based on Edge Intelligent Terminals, Chinese Journal of Ordnance Industry, February 26, 2023

Zhang Peixu, Hu Guanyu, Yang Xinyu, A Domain Enhanced and Adaptive Paradigm for Person Re-ID in Dress Change, Journal of Xidian University, August 2023

Zhang Mingzhen. Underground Pedestrian Detection Model Based on Dense-YOLO Network, Industrial and Mining Automation, 2022, 48(03)

Guan Jiacheng. Improved Lightweight Target Detection Based on YOLOv5, Computer System Applications, 2023, 32(09)




How to Cite

“Real-time Pedestrian Tracking Based on YOLOv3 and Prototype Clustering” (2023) Transactions on Engineering and Technology Research, 1, pp. 137–142. doi:10.62051/zhes5p18.

Similar Articles

1-10 of 12

You may also start an advanced similarity search for this article.