Adaptive Randomized Smoothing: Certified Adversarial Robustness for Multi-Step Defences
Saiyue Lyu, Shadab Shaikh, Frederick Shpilevskiy, Evan Shelhamer, Mathias Lécuyer
Abstract
We propose Adaptive Randomized Smoothing (ARS) to certify the predictions of our test-time adaptive models against adversarial examples. ARS extends the analysis of randomized smoothing using $f$-Differential Privacy to certify the adaptive composition of multiple steps. For the first time, our theory covers the sound adaptive composition of general and high-dimensional functions of noisy inputs. We instantiate ARS on deep image classification to certify predictions against adversarial examples of bounded $L_{\infty}$ norm. In the $L_{\infty}$ threat model, ARS enables flexible adaptation through high-dimensional input-dependent masking. We design adaptivity benchmarks, based on CIFAR-10 and CelebA, and show that ARS improves standard test accuracy by 1 to 15\% points. On ImageNet, ARS improves certified test accuracy by up to 1.6% points over standard RS without adaptivity. Our code is available at [https://github.com/ubc-systopia/adaptive-randomized-smoothing](https://github.com/ubc-systopia/adaptive-randomized-smoothing).
BibTeX
@inproceedings{
lyu2024adaptive,
title={Adaptive Randomized Smoothing: Certified Adversarial Robustness for Multi-Step Defences},
author={Saiyue Lyu and Shadab Shaikh and Frederick Shpilevskiy and Evan Shelhamer and Mathias L{\'e}cuyer},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=MN4nt01TeO}
}