Event
Proposing an Enumeration-Guided Training Framework for Neural HJ Safety Certificates
1 reports · 1 sources
The researchers propose a verification-guided training framework that uses probabilistic enumeration of safety constraints to detect forward invariance violations in neural Hamilton-Jacobi safety certificates, identifying states where the certificate incorrectly judges behavior as safe, and generating counterexample sets that are more compact and closer to the failure boundary than standard adversarial attack methods. After fine-tuning a pretrained HJ value function with these counterexamples, the certifiable region expands and certification time decreases on all tested benchmarks. The authors also introduce a deployment-time shielding mechanism based on the refined certificate; in Isaac Sim physics simulations and real autonomous racing experiments, it achieves zero safety constraint violations compared with a nominal controller and an LQR-based backup strategy.
Report timeline
- Research on Enumeration-Guided Training of Neural Hamilton-Jacobi Safety Certificates
This research proposes a verification-guided training framework that uses probabilistic enumeration of safety constraints to detect forward-invariance violations, generating counterexample sets that are more compact and closer to the failure boundary than those from standard adversarial attack methods. Fine-tuning a pretrained HJ value function with these counterexamples expands the certifiable region and shortens certification time across all tested benchmarks; the authors also introduce a deployment-time shielding mechanism based on the refined certificates, achieving zero safety constraint violations in Isaac Sim physics simulation and real-world autonomous racing experiments compared with a nominal controller and an LQR-based backup strategy.
arXiv cs.RO ·