Contractive Dynamical Imitation Policies for Efficient Out-of-Sample Recovery
This addresses the problem of unreliable policy deployment in robotics for more robust imitation learning, though it appears incremental as it builds on stable dynamical systems with a focus on transient behavior.
The paper tackles the problem of unreliable outcomes in out-of-sample regions for imitation learning by proposing a framework for learning policies modeled by contractive dynamical systems, which ensures convergence and enables efficient recovery, demonstrating substantial performance improvements in simulated robotic tasks.
Imitation learning is a data-driven approach to learning policies from expert behavior, but it is prone to unreliable outcomes in out-of-sample (OOS) regions. While previous research relying on stable dynamical systems guarantees convergence to a desired state, it often overlooks transient behavior. We propose a framework for learning policies modeled by contractive dynamical systems, ensuring that all policy rollouts converge regardless of perturbations, and in turn, enable efficient OOS recovery. By leveraging recurrent equilibrium networks and coupling layers, the policy structure guarantees contractivity for any parameter choice, which facilitates unconstrained optimization. We also provide theoretical upper bounds for worst-case and expected loss to rigorously establish the reliability of our method in deployment. Empirically, we demonstrate substantial OOS performance improvements for simulated robotic manipulation and navigation tasks.