Probabilistic Robustness Analysis in High Dimensional Space: Application to Semantic Segmentation Network
This work addresses robustness verification for safety-critical applications like medical imaging and autonomous driving, offering a scalable and less conservative method, though it is incremental as it builds on existing conformal inference techniques.
The paper tackles the problem of probabilistic robustness verification for semantic segmentation networks, which often fails to scale and yields overly conservative guarantees; the proposed framework, using conformal inference with a novel clipping block, provides reliable safety guarantees while substantially reducing conservatism across multiple large-scale datasets.
Semantic segmentation networks (SSNs) are central to safety-critical applications such as medical imaging and autonomous driving, where robustness under uncertainty is essential. However, existing probabilistic verification methods often fail to scale with the complexity and dimensionality of modern segmentation tasks, producing guarantees that are overly conservative and of limited practical value. We propose a probabilistic verification framework that is architecture-agnostic and scalable to high-dimensional input-output spaces. Our approach employs conformal inference (CI), enhanced by a novel technique that we call the \textbf{clipping block}, to provide provable guarantees while mitigating the excessive conservatism of prior methods. Experiments on large-scale segmentation models across CamVid, OCTA-500, Lung Segmentation, and Cityscapes demonstrate that our framework delivers reliable safety guarantees while substantially reducing conservatism compared to state-of-the-art approaches on segmentation tasks. We also provide a public GitHub repository (https://github.com/Navidhashemicodes/SSN_Reach_CLP_Surrogate) for this approach, to support reproducibility.