A Tool for Neural Network Global Robustness Certification and Training
This work addresses the need for scalable global robustness certification in safety-critical systems, representing an incremental advance over prior optimization-based approaches.
The paper tackles the problem of certifying global robustness for neural networks, which is limited to small networks by existing methods, and presents GROCET, a GPU-supported framework that improves efficiency and enables differentiable training for globally robust networks.
With the increment of interest in leveraging machine learning technology in safety-critical systems, the robustness of neural networks under external disturbance receives more and more concerns. Global robustness is a robustness property defined on the entire input domain. And a certified globally robust network can ensure its robustness on any possible network input. However, the state-of-the-art global robustness certification algorithm can only certify networks with at most several thousand neurons. In this paper, we propose the GPU-supported global robustness certification framework GROCET, which is more efficient than the previous optimization-based certification approach. Moreover, GROCET provides differentiable global robustness, which is leveraged in the training of globally robust neural networks.