Connecting Certified and Adversarial Training

Yuhao Mao, Mark Niklas Mueller, Marc Fischer, Martin Vechev

Keywords: Certified Training, Certified Robustness, Adversarial Robustness, Robustness Verification

TL;DR: We combine certified and adversarial training, reducing over-regularization, and thus leading to state-of-the-art performance in certified accuracy.

Abstract:

Training certifiably robust neural networks remains a notoriously hard problem.
While adversarial training optimizes under-approximations of the worst-case loss, which leads to insufficient regularization for certification, sound certified training methods, optimize loose over-approximations, leading to over-regularization and poor (standard) accuracy.
In this work, we propose TAPS, an (unsound) certified training method that combines IBP and PGD training to optimize more precise, although not necessarily sound, worst-case loss approximations, reducing over-regularization and increasing certified and standard accuracies.
Empirically, TAPS achieves a new state-of-the-art in many settings, e.g., reaching a certified accuracy of $22$% on TinyImageNet for $\ell_\infty$-perturbations with radius $\epsilon=1/255$. We make our implementation and networks public at https://github.com/eth-sri/taps.

Submission Number: 11952