A Simple Approach to Constraint-Aware Imitation Learning with Application to Autonomous Racing
This addresses safety and performance issues in autonomous racing, but it is incremental as it builds on existing imitation learning methods.
The paper tackled the problem of ensuring constraint satisfaction in imitation learning for high-precision tasks like autonomous racing, and the result was a simple method that improved constraint satisfaction and consistency compared to Behavior Cloning, as validated through simulations.
Guaranteeing constraint satisfaction is challenging in imitation learning (IL), particularly in tasks that require operating near a system's handling limits. Traditional IL methods, such as Behavior Cloning (BC), often struggle to enforce constraints, leading to suboptimal performance in high-precision tasks. In this paper, we present a simple approach to incorporating safety into the IL objective. Through simulations, we empirically validate our approach on an autonomous racing task with both full-state and image feedback, demonstrating improved constraint satisfaction and greater consistency in task performance compared to BC.