Fail-Safe Adversarial Generative Imitation Learning
This work addresses safety-critical imitation learning for applications like autonomous driving, though it appears incremental as it builds on existing adversarial and safety techniques.
The paper tackled the problem of ensuring safety in imitation learning by proposing a modular method with a safety layer that provides worst-case safety guarantees and closed-form probability densities for generative policies. In experiments on real-world driver interaction data, the method demonstrated tractability, safety, and imitation performance.
For flexible yet safe imitation learning (IL), we propose theory and a modular method, with a safety layer that enables a closed-form probability density/gradient of the safe generative continuous policy, end-to-end generative adversarial training, and worst-case safety guarantees. The safety layer maps all actions into a set of safe actions, and uses the change-of-variables formula plus additivity of measures for the density. The set of safe actions is inferred by first checking safety of a finite sample of actions via adversarial reachability analysis of fallback maneuvers, and then concluding on the safety of these actions' neighborhoods using, e.g., Lipschitz continuity. We provide theoretical analysis showing the robustness advantage of using the safety layer already during training (imitation error linear in the horizon) compared to only using it at test time (up to quadratic error). In an experiment on real-world driver interaction data, we empirically demonstrate tractability, safety and imitation performance of our approach.