This research aims to enhance the efficacy of object detection methodologies in complex traffic scenarios. The focus is on introducing an intelligent approach for identifying vehicles and pedestrians, leveraging the YOLOv5s framework. The study delves into the YOLOv5s architecture, implementing improvements by substituting the BottleNeck CSP module with the Ghost-BottleNeck module to decrease the computational burden. Additionally, the loss function is refined by incorporating Alpha-IoU as the localization loss function, effectively minimizing the discrepancy between predicted and actual bounding boxes. Furthermore, from an inference perspective, the SAHI strategy is introduced to enhance the inference process, significantly improving the detection performance of pedestrians. Finally, ablation experiments and real traffic road scene trials confirm that the SDYOLOv5 pedestrian detection method not only sustains commendable real-time performance but also elevates accuracy in pedestrian detection. This research holds significant implications for advancing intelligent vehicle testing and detection.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    A YOLOv5s-based Method for Intelligent Pedestrian Detection in Vehicles


    Contributors:
    Feng, Bin (author) / Wang, Zichen (author) / Bi, Wanglinfeng (author)


    Publication date :

    2024-05-17


    Size :

    516463 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English





    UAV target detection algorithm based on improved YOLOv5s

    Zhang, Tao / Wang, Fenmei / Chen, Dongxu et al. | SPIE | 2023


    Automatic Inspection Method for YOLOv5s Railway Freight Vehicles Based on Channel Pruning

    Wang, Shisheng / Du, Chengxin / Meng, Yukun et al. | IEEE | 2023


    Research on Vehicle Detection Method Based on Improved YOLOv5s

    Ma, Liangliang / Zhong, Runlu / Shi, Xiaohong et al. | IEEE | 2024