YOLOv8 Traffic Sign Detection Incorporating Attention Mechanisms

Aiming at the problems of existing traffic sign detection algorithms limited by light, occlusion and low detection accuracy, a YOLOv8 traffic sign detection algorithm incorporating an attention mechanism is proposed. By introducing the attention mechanism, the model can be allowed to automatically select the region of interest, thus reducing the unnecessary computational overhead and decreasing the influence of interference factors, thus improving the speed and accuracy of model detection. The experimental results show that the improved algorithm improves the accuracy and average accuracy by 2.9% and 2% respectively compared with the original algorithm, and the frame rate reaches 112FPS, which meets the requirements of real-time detection.