CCSPNet-Joint: Efficient Joint Training Method for Traffic Sign Detection Under Extreme Conditions

Traffic sign detection is an important research direction in intelligent driving. Unfortunately, existing methods often overlook training methods specifically designed for extreme conditions such as fog, rain, and motion blur. Moreover, the end-to-end training strategy for image denoising and object detection models fails to utilize inter-model information effectively. To address these issues, we propose CCSPNet, an efficient feature extraction module based on Contextual Transformer and CNN, capable of effectively utilizing the static and dynamic features of images, achieving faster inference speed and providing stronger feature enhancement capabilities. Furthermore, we establish the correlation between object detection and image denoising tasks and propose a joint training model, CCSPNet-Joint, to improve data efficiency and generalization. Finally, to validate our approach, we create the CCTSDB-AUG dataset for traffic sign detection in extreme scenarios. Extensive experiments have shown that CCSPNet achieves state-of-the-art performance in traffic sign detection under extreme conditions. Compared to end-to-end methods, CCSPNet-Joint achieves a 5.32% improvement in precision and an 18.09% improvement in mAP@.5.

I. Introduction

Traffic sign detection (TSD) plays a significant role in the field of intelligent driving by providing vital road information to intelligent driving systems, enabling accurate recognition for subsequent decision-making processes [7], [32]. Traffic sign detection algorithms utilize computer vision techniques to rapidly and accurately identify and extract information from images or video data pertaining to traffic signs [24]. The application of such algorithms aids intelligent vehicles in the real-time acquisition of road sign information, enhancing driving safety and overall driving efficiency.