Semantic segmentation neural networks (SSNs) are increasingly essential in highstakes fields such as medical imaging, autonomous driving, and environmental monitoring, where robustness to input uncertainties and adversarial examples is crucial for ensuring safety and reliability. However, traditional probabilistic verification methods struggle to scale effectively with the size and depth of modern SSNs, especially when dealing with their high-dimensional, structured inputs/outputs. As the output dimension increases, these methods tend to become overly conservative, resulting in unnecessarily restrictive safety guarantees. In this work, we propose a probabilistic, data-driven verification algorithm that is architecture-agnostic and scalable, capable of handling the high-dimensional outputs of SSNs without introducing conservative and loose guarantees. We leverage efficient sampling-based reachability analysis to explore the space of possible outputs while maintaining computational feasibility.

A neural ordinary differential equations network (ODE-Net)-enabled reachability method (Neuro-Reachability) is devised for the dynamic verification of networked microgrids (NMs) with unidentified subsystems and heterogeneous uncertainties. Three new contributions are presented: 1) An ODENet-enabled dynamic model discovery approach is devised to construct the data-driven state-space model which preserves the nonlinear and differential structure of the NMs system; 2) A physics-data-integrated (PDI) NMs model is established, which empowers various NM analytics; and 3) A conformance-empowered reachability analysis is developed to enhance the reliability of the PDI-driven dynamic verification. Extensive case studies demonstrate the efficacy of the ODE-Net-enabled method in microgrid dynamic model discovery, and the effectiveness of the Neuro-Reachability approach in verifying the NMs dynamics under multiple uncertainties and various operational scenarios.