Recognition of Traffic Signs using mobile-net neural network
Paper ID : 1114-MVIP2020
mohamadhadi safi *1, Roozbeh Rajabi2
1student, Qom university of technology
Traffic signs detection has many applications, including in self-driving cars, traffic mapping and accident reduction.More recently, deep learning models have been used to detect traffic signs as a way of finding effective features and also classifying them.In this paper, using Convolution Neural Network (CNN), MobileNetV2 which is a network with lower number of parameters and less weight than other networks including AlexNet and GoogleNet and set Traffic Sign Data (GTSRB) As a database, a method for detecting traffic signs is provided. The network achieved 99.90% learning accuracy and 99.41% recognition accuracy.
Recognition, Traffic Signs, Mobile-net, Deep Learning
Status : Paper Accepted (Oral Presentation)