Certified geometric robustness -- Super-DeepG
For safety-critical applications requiring certified robustness to geometric image perturbations, Super-DeepG provides a more precise and efficient verification tool.
This paper addresses formal verification of neural networks against geometric perturbations (rotation, scaling, shearing, translation). Super-DeepG improves linear relaxation and Lipschitz optimization, achieving better precision and computational efficiency than prior work.
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.