Spatiotemporal Costmap Inference for MPC via Deep Inverse Reinforcement Learning
This work addresses the challenge of making autonomous vehicles behave naturally in traffic, which is an incremental improvement over existing methods.
The paper tackled the problem of autonomously producing natural driver behavior by learning a reward function from human demonstrations, resulting in higher success rates for autonomous driving tasks in dense traffic compared to baseline methods.
It can be difficult to autonomously produce driver behavior so that it appears natural to other traffic participants. Through Inverse Reinforcement Learning (IRL), we can automate this process by learning the underlying reward function from human demonstrations. We propose a new IRL algorithm that learns a goal-conditioned spatiotemporal reward function. The resulting costmap is used by Model Predictive Controllers (MPCs) to perform a task without any hand-designing or hand-tuning of the cost function. We evaluate our proposed Goal-conditioned SpatioTemporal Zeroing Maximum Entropy Deep IRL (GSTZ)-MEDIRL framework together with MPC in the CARLA simulator for autonomous driving, lane keeping, and lane changing tasks in a challenging dense traffic highway scenario. Our proposed methods show higher success rates compared to other baseline methods including behavior cloning, state-of-the-art RL policies, and MPC with a learning-based behavior prediction model.