In recent years, unmanned systems have become an increasingly popular research direction, in which the fast and accurate detection of traffic signs is an important research element in unmanned systems. In order to apply the target detection algorithm to embedded devices and mobile devices with low computing power, a G-GhostNet-YOLOx algorithm is proposed, which is based on the YOLOx-m network structure with the introduction of G-GhostNet lightweight network. The model is made as small as possible with the premise of ensuring the accuracy of the model. The experimental results show that the mAP (Mean Average Precision) of the Chinese traffic sign test set reaches 93.4%, which is high and small enough for application in embedded and mobile devices, which will be of practical significance for promoting the development of driverless technology and improving traffic safety.
Traffic sign detection research based on G-GhostNet-YOLOx model
2022-10-12
1452323 byte
Conference paper
Electronic Resource
English
Research on YOLOX-based tire defect detection method
IEEE | 2022
|Residual YOLOX-based Ship Object Detection Method
IEEE | 2022
|YOLOX-based dynamic vehicle target detection
IEEE | 2022
|