A traffic sign detection model based on coordinate attention - bidirectional feature pyramid network

A traffic sign detection model based on coordinate attention - bidirectional feature pyramid network

Blog Article

ultra max dog shampoo In the field of autonomous driving, the correct detection of traffic signs can provide important information for environmental perception.To address the low recognition rate and misdetection and missed detection issues of various traffic signs, we propose a coordinate attention - bidirectional feature pyramid network (CA-BIFPN) traffic sign detection model.This model combines a you only look once version 5 (YOLOv5) network and a coordinate attention (CA) mechanism, and introduces a bidirectional feature pyramid network (BIFPN) with skip-connection feature fusion to improve the utilization efficiency of multi-scale semantic features, and to enhance the detection accuracy of traffic signs while improving the detection efficiency of small target objects.We conducted experiments tumbler lid that holds liquor using the traffic sign data set TT100K as the test object to compare the detection accuracy of the CA-BIFPN model with that of the single shot multibox detector (SSD) and YOLOv5 model.

The detection accuracy of the CA-BIFPN traffic sign detection model are improved by 4.5% and 1.3% compared with the SSD model and YOLOv5 model, respectively, validating the effectiveness of the proposed model.