Article Details

Title · Authors · Abstract · Keywords · DOI · References

Research on Pedestrian Detection System Based on YOLOv5

JIAO Tianwen, TIAN Xiuyun

(School of Electronics and Information Engineering, Guangdong Ocean University, Zhanjiang 524088, China)

Section：Computer Technology · Vol. 2025, Issue 2 · Pages：33-37,45 · 2096-4706（2025）02-0033-06 · CLC number：TP391.4 · Document code：A · DOI：10.19850/j.cnki.2096-4706.2025.02.006

Abstract

Aiming at problems of the amount of calculation with redundancy and low accuracy, an improved YOLOv5 pedestrian detection model is proposed. This system uses Ghost convolution to combine with the Transformer Self-Attention Mechanism, then combines with the BiFPN structure and EIoU loss function. The INRIA pedestrian detection data set is divided into the training set, the validation set, and the test set according to the ratio of 7∶2∶1. The SGD optimizer is used to train the model for 300 Epochs, and the trained weight model is used to test the test set. The results show that the average accuracy of the improved model detection is increased by 1.5%, and the amount of calculation is significantly reduced.

Keywords

pedestrian detection; YOLOv5 model; Ghost convolution; BiFPN

References

  1. [1] Su Songzhi, Li Shaozi, Chen Shuyuan, etc Overview of Pedestrian Detection Technology [J]. Acta Sinica, 2012, 40 (4): 814-820
  2. [2] DALAL N，TRIGGS B. Histograms of Oriented Gradients for Human Detection [C]//2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).San Diego：IEEE，2005，1：886-893.
  3. [3] VIOLA P，JONES M. Rapid Object Detection Using a Boosted Cascade of Simple Features [C]//Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2001).Kauai：IEEE，2001：I-I.
  4. [4] REDMON J，DIVVALA S，GIRSHICK R，et al. You Only Look Once: Unified, Real-Time Object Detection [C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Las Vegas：IEEE，2016：779-788.
  5. [5] HE K M，ZHANG X Y，REN S Q，et al. Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition [J].IEEE Transactions on Pattern Analysis and Machine Intelligence，2015，37（9）：1904-1916.
  6. [6] HAN K，WANG Y H，TIAN Q，et al. GhostNet: More Features From Cheap Operations [C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).Seattle：IEEE，2020，1577-1586.
  7. [7] Peng Bin, Bai Jing, Li Wenjing, etc Research progress on visual transformers for image classification [J]. Computer Science and Exploration, 2024, 18 (2): 320-344
  8. [8] TAN M X，PANG R，LE Q V. EfficientDet: Scalable and Efficient Object Detection [C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).Seattle：IEEE，2020：10778-10787.
  9. [9] YU C S，SHIN Y.SAR Ship Detection Based On Improved YOLOv5 and BiFPN [J].ICT Express，2024，10（1）：28-33.
  10. [10] ZHANG Y F，REN W Q，ZHANG Z，et al. Focal and Efficient IOU Loss for Accurate Bounding Box Regression [J].Neurocomputing，2022，506：146-157.

Funding

2022 Guangdong Ocean University Campus Course - Mathematical Physics Methods (010301112202)

About the authors

Jiao Tianwen (2001-), male, Han ethnicity, from Zhuhai, Guangdong, undergraduate degree, research direction: embedded systems, AI algorithms; Corresponding author: Tian Xiuyun (1974-), female, Han ethnicity, from Zhanjiang, Guangdong, lecturer, master's degree, research direction: software algorithms, optoelectronic detection.