YOLO-AEF: traffic sign detection on challenging traffic scenes via adaptive enhancement and fusion

With the rapid development of intelligent driving technologies such as autonomous driving and driver assistance systems, traffic sign detection has become increasingly vital for road and transportation safety. To address performance degradation under complex illumination conditions, this paper proposes YOLO with Adaptive Enhancement Fusion (YOLO_AEF), a novel framework that integrates adaptive enhancement and dual-branch feature fusion. The Multiple Exposure Enhancement Module (MEEM) adaptively enhances low-quality images under challenging illumination conditions, while the Adaptive Feature Fusion Module (AFFM) fuses semantic features from both the original and enhanced images, improving robustness and contextual representation. The Fusion Detection Module (FDM) realizes robust detection of traffic signs following the framework of YOLO. Experimental results show that YOLO_AEF outperforms DEIM on the TT100K dataset by 4.8 percentage points and surpasses YOLOv12 on the CCTSDB dataset by 7.0 percentage points demonstrating strong generalization capability. Compared with various state-of-the-art methods, YOLO_AEF delivers superior accuracy and speed under diverse illumination challenges, making it highly suitable for intelligent traffic perception systems.

Text 124336.pdf - Published Version Restricted to Repository staff only Download (3MB) Official URL or Download Paper: https://www.sciencedirect.com/science/article/pii/...

Actions (login required)

View Item