This project presents methods for training Autonomous Driving Models using NeuroEvolution of Augmenting Topologies and Deep Learning and their implementation in 2D and 3D simulations.
#1 - Self-driving 2D
Training an autopilot model using genetic algorithm for the generation of evolving artificial neural networks on randomly generated highways in a 2D simulation.
#2 - Self-parking 2D
Training an autopilot model using genetic algorithm for the generation of evolving artificial neural networks on randomly generated parking lot in a 2D simulation.
#3 - Self-driving 3D
Training an autopilot deep learning model based on behavioral cloning metodology using Udacity's self driving car 3D simulator.