• Implemented a DDPG-based RL agent that learned to simulate realistic steering responses for a car given a desired trajectory. The implementation won the comma.ai challenge.

Screenshots: