Efficient randomized smoothing by denoising with learned score function

Kyungmin Lee, Seyoon Oh

Keywords: Adversarial Robustness, Provable Adversarial Defense, Randomized Smoothing, Image Denoising, Score Estimation

Abstract: The randomized smoothing with various noise distributions is a promising approach to protect classifiers from $\ell_p$ adversarial attacks. However, it requires an ensemble of classifiers trained with different noise types and magnitudes, which is computationally expensive. In this work, we present an efficient method for randomized smoothing that does not require any re-training of classifiers. We built upon denoised smoothing, which prepends denoiser to the pre-trained classifier. We investigate two approaches to the image denoising problem for randomized smoothing and show that using the score function suits for both. Moreover, we present an efficient algorithm that can scale to randomized smoothing and can be applied regardless of noise types or levels. To validate, we demonstrate the effectiveness of our methods through extensive experiments on CIFAR-10 and ImageNet, under various $\ell_p$ adversaries.

One-sentence Summary: We provide efficient method to generate smoothed classifier for provable defense by exploiting score-based image denoiser

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Reviewed Version (pdf): https://openreview.net/references/pdf?id=Sgdif_fK18