Traffic Sign Detection Model Based on Improved YOLOv5
To address the issues of low accuracy and poor recognition performance of traffic sign detection models in natural scenes,YOLOv5 is used as the basic network for improvement.Firstly,a convolutional attention module is introduced to obtain more detailed features.Secondly,BiFPN is used as the feature fusion network in the Neck layer to better achieve multi-scale feature fusion.Finally,deep hyperparametric convolution DO Conv is used instead of some traditional convolutions in the network to increase the number of learnable parameters.The results show that compared with the original YOLOv5,the improved YOLOv5 network model has improved detection accuracy by 1.2%and detection speed by 1.4 fps.This model can recognize traffic signs more quickly and accurately.
NETL
NSTL
万方数据
