English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

The current state-of-the-art defense methods against adversarial examplestypically focus on improving either empirical or certified robustness. Amongthem, adversarially trained (AT) models produce empirical state-of-the-artdefense against adversarial examples without providing any robustnessguarantees for large classifiers or higher-dimensional inputs. In contrast,existing randomized smoothing based models achieve state-of-the-art certifiedrobustness while significantly degrading the empirical robustness againstadversarial examples. In this paper, we propose a novel method, called\emph{Certification through Adaptation}, that transforms an AT model into arandomized smoothing classifier during inference to provide certifiedrobustness for $\ell_2$ norm without affecting their empirical robustnessagainst adversarial attacks. We also propose \emph{Auto-Noise} technique thatefficiently approximates the appropriate noise levels to flexibly certify thetest examples using randomized smoothing technique. Our proposed\emph{Certification through Adaptation} with \emph{Auto-Noise} techniqueachieves an \textit{average certified radius (ACR) scores} up to $1.102$ and$1.148$ respectively for CIFAR-10 and ImageNet datasets using AT models withoutaffecting their empirical robustness or benign accuracy. Therefore, our paperis a step towards bridging the gap between the empirical and certifiedrobustness against adversarial examples by achieving both using the sameclassifier.

Towards Bridging the gap between Empirical and Certified Robustness against Adversarial Examples