Teach a convolutional neural network (NVIDIA architecture) how to drive using the Udacity self-driving car simulator
-
Create an Anaconda environment using conda env create -f environment.yml --name car_environment within the repo
-
Activate the Anaconda environment using source activate car_environment
-
model.py --> script used to create and train the model
-
drive.py --> script to drive the car by Udacity
-
model.h5 --> a trained Keras model
-
video.py --> script used to create a video from the autonomous driving pictures
Save a video of the autonomous agent --> python drive.py model.h5 run1 (run1 being the directory where the images get saved)
Make an mp4 video of the autonomous driving --> python video.py run1 (--fps 48 for 48 FPS)
Starter code provided by Udacity