English

Certified geometric robustness -- Super-DeepG

Artificial Intelligence 2026-04-28 v1 Machine Learning Symbolic Computation

Abstract

Safety-critical applications are required to perform as expected in normal operations. Image processing functions are often required to be insensitive to small geometric perturbations such as rotation, scaling, shearing or translation. This paper addresses the formal verification of neural networks against geometric perturbations on their image dataset. Our method Super-DeepG improves the reasoning used in linear relaxation techniques and Lipschitz optimization, and provides an implementation that leverages GPU hardware. By doing so, Super-DeepG achieves both precision and computational efficiency of robustness certification, to an extent that outperforms prior work. Super-DeepG is shared as an open-source tool on GitHub.

Keywords

Cite

@article{arxiv.2604.24379,
  title  = {Certified geometric robustness -- Super-DeepG},
  author = {Noémie Cohen and Mélanie Ducoffe and Christophe Gabreau and Claire Pagetti and Xavier Pucel},
  journal= {arXiv preprint arXiv:2604.24379},
  year   = {2026}
}

Comments

ICCPS / HSCC 2026, CPS IoT Week, May 2026, Saint Malo (Palais du Grand Large), France

R2 v1 2026-07-01T12:37:04.882Z