In this paper, the application of the target pedestrian detection algorithm is studied based on the YOLOv4 network. The simulative results show that the improved YOLOv4 algorithm, which integrates the CBAM attention mechanism and Focal Loss function, shows high detection accuracy in target pedestrian detection. By CBAM, The new feature map will get the attention weight of the channel and space dimension, practical features of the target. Besides, focal loss enhanced the training of difficult-to-classify samples. Based on the improved YOLOv4 algorithm, which will have great potential, the problems of low detection accuracy and seriously missed pedestrian detection in realistic, complex visual scenes are solved.
Detection Algorithm Based on Improved YOLOv4 for Pedestrian
12.10.2022
1432056 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Pedestrian Detection Algorithm Based on Improved YOLOv4
IEEE | 2023
|Improved YOLOv4-tiny network for pedestrian detection
IEEE | 2022
|A Improved Yolov4’s vehicle and pedestrian detection method
VDE-Verlag | 2022
|UAV Target Detection Algorithm Based on Improved YOLOv4
Springer Verlag | 2022
|