Spatiotemporal Costmap Inference for MPC via Deep Inverse Reinforcement Learning

TL;DR

This paper introduces a novel deep IRL algorithm that learns goal-conditioned spatiotemporal reward functions, enabling MPC to perform autonomous driving tasks more effectively without manual cost function tuning.

Contribution

The paper presents GSTZ-MEDIRL, a new IRL method for learning reward functions that improve autonomous driving performance in dense traffic scenarios.

Findings

Higher success rates than baseline methods

Effective in dense highway traffic scenarios

Outperforms behavior cloning and RL policies

Abstract

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.