/traffic-sign-classification

Traffic sign classification using neural networks

Primary LanguageJupyter Notebook

Traffic Sign Classification with Neural Networks

The aim of this project is to implement and train several neural networks in PyTorch for the task of traffic sign classification. The dataset used is from the German Traffic Sign Recognition Benchmark (GTSRB) (reference below). Results and a more detailed description will be published in the future as this project is still WIP.

image

Architectures

Below is a list that tracks which of the planned architectures have already been implemented and trained. See networks.py for the implementation.

  • Single linear layer
  • Fully-connected network: 2 fully-connected layers
  • Convolutional network: 2 convolutional layers with max pooling, 2 fully-connected layers
  • Convolutional network with residual connections
  • (Variational) Autoencoder

References

J. Stallkamp, M. Schlipsing, J. Salmen, and C. Igel. The German Traffic Sign Recognition Benchmark: A multi-class classification competition. In Proceedings of the IEEE International Joint Conference on Neural Networks, pages 1453–1460. 2011.