DeepThin: A novel lightweight CNN architecture for traffic sign recognition without GPU requirements

作者：

Highlights：

• A novel CNN architecture is proposed designed for traffic sign recognition.

• The proposed CNN architecture is capable of first training without using GPU.

• Overlapping max pooling and sparsely strided convolution used for generalization.

• Proposed architecture is capable of beating human level performance.

摘要

•A novel CNN architecture is proposed designed for traffic sign recognition.•The proposed CNN architecture is capable of first training without using GPU.•Overlapping max pooling and sparsely strided convolution used for generalization.•Proposed architecture is capable of beating human level performance.

论文关键词：Traffic signs recognition,Convolutional Neural Network,Deep Learning,Image Augmentation,DeepThin CNN,Image classification,Ensemble learning

论文评审过程：Received 21 August 2020, Revised 15 November 2020, Accepted 7 December 2020, Available online 9 December 2020, Version of Record 15 December 2020.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.114481