In-Distribution Barrier Functions: Self-Supervised Policy Filters that Avoid Out-of-Distribution States
This addresses safety concerns in learning-based control for robotics, particularly for systems relying on high-dimensional visual observations, though it is incremental as it adapts existing control barrier functions to visual domains.
The paper tackles the problem of learned controllers behaving unsafely when encountering out-of-distribution states by proposing a control filter that encourages systems to stay in-distribution using offline safe demonstrations, demonstrating effectiveness in two visuomotor control tasks in simulation.
Learning-based control approaches have shown great promise in performing complex tasks directly from high-dimensional perception data for real robotic systems. Nonetheless, the learned controllers can behave unexpectedly if the trajectories of the system divert from the training data distribution, which can compromise safety. In this work, we propose a control filter that wraps any reference policy and effectively encourages the system to stay in-distribution with respect to offline-collected safe demonstrations. Our methodology is inspired by Control Barrier Functions (CBFs), which are model-based tools from the nonlinear control literature that can be used to construct minimally invasive safe policy filters. While existing methods based on CBFs require a known low-dimensional state representation, our proposed approach is directly applicable to systems that rely solely on high-dimensional visual observations by learning in a latent state-space. We demonstrate that our method is effective for two different visuomotor control tasks in simulation environments, including both top-down and egocentric view settings.