Learning Vision-Based Neural Network Controllers with Semi-Probabilistic Safety Guarantees
This addresses the critical problem of safety in autonomous systems with vision-based control, providing a novel framework that offers formal guarantees, though it is incremental in combining existing techniques like reachability analysis and generative networks.
The paper tackles the challenge of ensuring safety in vision-based autonomous systems by introducing a semi-probabilistic verification framework and a gradient-based training approach, achieving formal safety guarantees while maintaining strong nominal performance across multiple simulation and physical environments.
Ensuring safety in autonomous systems with vision-based control remains a critical challenge due to the high dimensionality of image inputs and the fact that the relationship between true system state and its visual manifestation is unknown. Existing methods for learning-based control in such settings typically lack formal safety guarantees. To address this challenge, we introduce a novel semi-probabilistic verification framework that integrates reachability analysis with conditional generative networks and distribution-free tail bounds to enable efficient and scalable verification of vision-based neural network controllers. Next, we develop a gradient-based training approach that employs a novel safety loss function, safety-aware data-sampling strategy to efficiently select and store critical training examples, and curriculum learning, to efficiently synthesize safe controllers in the semi-probabilistic framework. Empirical evaluations in X-Plane 11 airplane landing simulation, CARLA-simulated autonomous lane following, F1Tenth vehicle lane following in a physical visually-rich miniature environment, and Airsim-simulated drone navigation and obstacle avoidance demonstrate the effectiveness of our method in achieving formal safety guarantees while maintaining strong nominal performance.