Users could tailor deep learning models with architecture of their preference to train a car to self drive using this small scale simulation. There is also the option to drive the car manually .
Parameters such as lane count, and traffic and those pertaining to that of neural network such as training set size, number of sensors to detect collission, number of hidden layers and the corresponding nodes in it could be altered and experimented upon.
