TSD‐YOLO: Small traffic sign detection based on improved YOLO v8
Songjie Du,Weiguo Pan,Nuoya Li,Songyin Dai,Bingxin Xu,Hongzhe Liu,Cheng Xu,Xuewei Li
DOI: https://doi.org/10.1049/ipr2.13141
IF: 2.3
2024-06-04
IET Image Processing
Abstract:To address these issues and enhance detector performance, this paper proposes an improved algorithm based on YOLOv8s. To preserve fine‐grained information and improve small target detection, the authors incorporate the Space‐to‐Depth module into the sampling process. Additionally, the authors introduce Select Kernel attention mechanism, enabling the model to adaptively adjust the size of receptive field based on the target's dimensions, thereby focusing more effectively on the target area; Furthermore, the authors replace the original Complete Intersection over Union (CIoU) loss function with WIoUv3, enhancing the model's ability to fit the target box, accelerating convergence, overall improving detection performance. Traffic sign detection is critical for autonomous driving technology. However, accurately detecting traffic signs in complex traffic environments remains challenge despite the widespread use of one‐stage detection algorithms known for their real‐time processing capabilities. In this paper, the authors propose a traffic sign detection method based on YOLO v8. Specifically, this study introduces the Space‐to‐Depth (SPD) module to address missed detections caused by multi‐scale variations of traffic signs in traffic scenes. The SPD module compresses spatial information into depth channels, expanding the receptive field and enhancing the detection capabilities for objects of varying sizes. Furthermore, to address missed detections caused by complex backgrounds such as trees, this paper employs the Select Kernel attention mechanism. This mechanism enables the model to dynamically adjust its focus and more effectively concentrate on key features. Additionally, considering the uneven distribution of training data, the authors adopted the WIoUv3 loss function, which optimizes loss calculation through a weighted approach, thereby improving the model's detection performance across various sizes and frequencies of instances. The proposed methods were validated on the CCTSDB and TT100K datasets. Experimental results demonstrate that the authors' method achieves substantial improvements of 3.2% and 5.1% on the mAP50 metric compared to YOLOv8s, while maintaining high detection speed, significantly enhancing the overall performance of the detection system. The code for this paper is located at https://github.com/dusongjie/TSD‐YOLO‐Small‐Traffic‐Sign‐Detection‐Based‐on‐Improved‐YOLO‐v8
computer science, artificial intelligence,engineering, electrical & electronic,imaging science & photographic technology