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/

Zendy Plus

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

Zendy Tools

Zendy Open

Adversarial attacks rely on transferability, where an adversarial example(AE) crafted on a surrogate classifier tends to mislead a target classifier.Recent ensemble methods demonstrate that AEs are less likely to misleadmultiple classifiers in an ensemble. This paper proposes a new ensembletraining using a Pairwise Adversarially Robust Loss (PARL) that by constructionproduces an ensemble of classifiers with diverse decision boundaries. PARLutilizes outputs and gradients of each layer with respect to network parametersin every classifier within the ensemble simultaneously. PARL is demonstrated toachieve higher robustness against black-box transfer attacks than previousensemble methods as well as adversarial training without adversely affectingclean example accuracy. Extensive experiments using standard Resnet20,WideResnet28-10 classifiers demonstrate the robustness of PARL againststate-of-the-art adversarial attacks. While maintaining similar clean accuracyand lesser training time, the proposed architecture has a 24.8% increase inrobust accuracy ($\epsilon$ = 0.07) from the state-of-the art method.

Guardian of the Ensembles: Introducing Pairwise Adversarially Robust Loss for Resisting Adversarial Attacks in DNN Ensembles