Improving YOLOv8 via Triple Attention for Pedestrian Detection

Pedestrian detection, as an essential component of autonomous driving tasks, directly impacts the safety of vehicle operation and remains a hot and challenging research area in computer vision. In order to enhance the detection performance of pedestrian detection, this paper proposes an improved algorithm based on the YOLOv8 framework, named YoloV8-TA, by modifying the feature extraction network of the YOLOv8 algorithm and integrating the Triple Attention mechanism to enhance feature extraction capability. Experimental results demonstrate that the proposed method achieves improvements in accuracy, recall and mAP on the MoT17 dataset, with values of 82.67, 77.14 and 85.04, respectively, compared to the original YOLOv8, which represents enhancements of 2.05, 0.41 and 0.48,respectively.