Traffic Sign Detection and Recognition under Foggy Conditions
Ming Zhang, Cheng Li, Dequ Chen, Hui Zeng · 2023
As one of the key technologies of autonomous driving, traffic sign detection and recognition (TSDR) is still a challenging task due to the difficulty and complexity of traffic signs. In addition, the real world often has bad weather conditions, which makes detection more difficult. In this paper, we propose a multi-source traffic sign dataset for foggy scenes, named Foggy-DFG, which contains a large number of street view data for foggy conditions. A YOLOX-CBAM model with attention module is proposed. The attention module improves the performance of YOLOX by fusing and emphasizing the attention of specific locations to emphasize traffic signs under foggy conditions. Experiments on Foggy Driving dataset shows that the proposed method has achieved significant improvement.