Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Apr 30:10:e2021.
doi: 10.7717/peerj-cs.2021. eCollection 2024.

Research on multi-object detection technology for road scenes based on SDG-YOLOv5

Zhenyang Lv  1 , Rugang Wang  1 , Yuanyuan Wang  1 , Feng Zhou  1 , Naihong Guo  2
Affiliations

Research on multi-object detection technology for road scenes based on SDG-YOLOv5

Zhenyang Lv et al. PeerJ Comput Sci. .

Abstract

To resolve the challenges of low detection accuracy and inadequate real-time performance in road scene detection, this article introduces the enhanced algorithm SDG-YOLOv5. The algorithm incorporates the SIoU Loss function to accurately predict the angle loss of bounding boxes, ensuring their directionality during regression and improving both regression accuracy and convergence speed. A novel lightweight decoupled heads (DHs) approach is employed to separate the classification and regression tasks, thereby avoiding conflicts between their focus areas. Moreover, the Global Attention Mechanism Group Convolution (GAMGC), a lightweight strategy, is utilized to enhance the network's capability to process additional contextual information, thereby improving the detection of small targets. Extensive experimental analysis on datasets from Udacity Self Driving Car, BDD100K, and KITTI demonstrates that the proposed algorithm achieves improvements in mAP@.5 of 2.2%, 3.4%, and 1.0% over the original YOLOv5, with a detection speed of 30.3 FPS. These results illustrate that the SDG-YOLOv5 algorithm effectively addresses both detection accuracy and real-time performance in road scene detection.

Keywords: Attention mechanism; Decoupled detection head; Intelligent driving; Road scene detection.

PubMed Disclaimer

Conflict of interest statement

Naihong Guo is employed by Yancheng XiongYing Precision Machinery Company Limited.

Figures

Figure 1
Figure 1. SDG-YOLOv5 network structure.
Figure 2
Figure 2. The concepts of SIoU Loss regression.
Figure 3
Figure 3. Decoupled heads diagram.
Figure 4
Figure 4. Structure of the GAM network.
Figure 5
Figure 5. Channel attention module.
Figure 6
Figure 6. Spatial attention module.
Figure 7
Figure 7. Improved attention mechanism structure diagram.
Figure 8
Figure 8. Sample images extracted from datasets.
Figure 9
Figure 9. Accuracy chart before and after training on Udacity Self Driving Car dataset.
Figure 10
Figure 10. Comparison of test results obtained from different algorithms.
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Bochkovskiy A, Wang C-Y, Liao H-Y. YOLOv4: optimal speed and accuracy of object detection. ArXiv. 2020. - DOI
    1. Cai Y, Luan T, Gao H, Wang H, Chen L, Li Y, Sotelo M, Li Z. YOLOv4-5D: an effective and efficient object detector for autonomous driving. IEEE Transactions on Instrumentation and Measurement. 2021;70:4503613. doi: 10.1109/TIM.2021.3065438. - DOI
    1. Du Y, Qin B, Zhao C, Zhu Y, Cao J, Ji Y. A novel spatio-temporal synchronization method of roadside asynchronous MMW radar-camera for sensor fusion. IEEE Transactions on Intelligent Transportation Systems. 2022;23:22278–22289. doi: 10.1109/TITS.2021.3119079. - DOI
    1. Ge Z, Liu S, Wang F, Li Z, Sun J. YOLOX: exceeding YOLO series in 2021. ArXiv. 2021. - DOI
    1. Girshick R, Donahue J, Darrell T, Malik J. Rich feature hierarchies for accurate object detection and semantic segmentation. 2014 IEEE conference on computer vision and pattern recognition, Columbus, OH, USA, pp; Piscataway. 2014. pp. 580–587. - DOI

LinkOut - more resources