- The end goal of the project is to classify different traffic signs correctly.
- The dataset used is the project is GTSRB - German Traffic Sign Recognition Benchmark
- There are 43 different classes present in the dataset.
- Before running the .ipynb file, upload the
signnames.csv
and yourkaggle.json
credential files either directly to the environment or pull it using google colab. - Model predictions can be found in the
Prediction.ipynb
notebook or in the example_output directory. - The trained model is saved in the saved_model directory.
- This project is based on the tutoral blog written by Dr. Adrian Rosebrok -> Traffic Sign Classification with Keras and Deep Learning