Project: Build a Traffic Sign Recognition Program

This project is an implementation of Deep Neural Networks and Image processing to Classify Traffic Signs in the German Traffic Sign Dataset.

Dependencies

• Python 3.5
• TensorFlow 1.4
• Jupyter
• NumPy
• SciPy
• Matplotlib
• Open CV2

Dataset

1. Download the dataset.
2. This is a pickled dataset which has already resized the images to 32x32.

Steps

1. Clone the project and start the notebook.

git clone https://github.com/Mohit-Ak/CarND-Traffic-Sign-Classifier.git
cd CarND-Traffic-Sign-Classifier
jupyter notebook Traffic_Sign_Classifier.ipynb

2. Follow the instructions in the Traffic_Signs_Recognition.ipynb notebook.

FINAL ACCURACY

Dataset	Accuracy
Validation	94.2%
Test	93.1%

AUTHOR

Mohit Arvind Khakharia