A DDPG-based lateral controller trained with online MPC-PID demonstrations tracks paths in CARLA better than its demonstrator and tolerates noisy waypoints.
Improved Deep Reinforcement Learning with Expert Demonstrations for Urban Autonomous Driv- ing
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Autonomous Vehicle Lateral Control Using Deep Reinforcement Learning with MPC-PID Demonstration
A DDPG-based lateral controller trained with online MPC-PID demonstrations tracks paths in CARLA better than its demonstrator and tolerates noisy waypoints.