Self-Driving Car Engineer Nanodegree

Deep Learning

Project: Build a Traffic Sign Recognition Program

Overview

In this project, you will use what you've learned about deep neural networks and convolutional neural networks to classify traffic signs. You will train a model so it can decode traffic signs from natural images by using the German Traffic Sign Dataset. After the model is trained, you will then test your model program on new images of traffic signs you find on the web, or, if you're feeling adventurous pictures of traffic signs you find locally!

Dependencies

This project requires Python 3.5 and the following Python libraries installed:

• Jupyter
• NumPy
• SciPy
• scikit-learn
• TensorFlow

Setup

OS X and Linux

Install Anaconda

This project requires Anaconda and Python 3.4 or higher. If you don't meet all of these requirements, install the appropriate package(s).

Run the Anaconda Environment

Run these commands in your terminal to install all requirements:

$ git clone https://github.com/matthewzimmer/traffic-sign-classification.git
$ conda env create -f environment.yml

Install Tensorflow

GPU

AWS GPU Configuration

SSH instructions

Instance ID i-aab307f5 ssh to box aws ec2 get-console-output --instance-id i-aab307f5

Local

$ source activate traffic-sign-classification
$ export TF_BINARY_URL=https://storage.googleapis.com/tensorflow/mac/gpu/tensorflow-0.11.0-py3-none-any.whl
$ pip3 install --ignore-installed --upgrade $TF_BINARY_URL

NOTE: This is still buggy. When I try loading tensorflow, it errors out saying it can't find libcudart.8.0.dylib. Problem is, I have CUDA 7.5 installed.

CPU

$ conda install --name traffic-sign-classification -c conda-forge tensorflow

Run this command at the terminal prompt to install OpenCV. Useful for image processing:

$ conda install -c https://conda.anaconda.org/menpo opencv3

Run the Notebook

Start a Jupyter Server

Make sure to run the server from the same directory that you ran in the Setup steps above.

OS X and Linux

$ source activate traffic-sign-classification
$ jupyter notebook

Useful Conda Commands

Update a conda environment

$ conda env update -f environment.yml

Dataset

1. Download the dataset. You can download the pickled dataset in which we've already resized the images to 32x32 here.

2. Clone the project and start the notebook.

git clone https://github.com/udacity/traffic-signs
cd traffic-signs
jupyter notebook Traffic_Signs_Recognition.ipynb

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