We trained a convolutional neural network (CNN) to map raw pixels from three cameras ( Right, Center
and Left) directly to steering commands. This end-to-end approach proved surprisingly powerful.
With minimum training data from humans the system learns to drive on roads with or without lane
markings. The system automatically learns internal representations of the necessary processing steps
such as detecting useful road features with only the human steering angle as the training signal. We
never explicitly trained it to detect, for example, the outline of roads.
The Udacity simulator captures images at 24 frames per second (fps) from three cameras (right, center,
and left) and stores them as JPG images.
