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

Regional adversarial attacks often rely on complicated methods for generatingadversarial perturbations, making it hard to compare their efficacy againstwell-known attacks. In this study, we show that effective regionalperturbations can be generated without resorting to complex methods. We developa very simple regional adversarial perturbation attack method usingcross-entropy sign, one of the most commonly used losses in adversarial machinelearning. Our experiments on ImageNet with multiple models reveal that, onaverage, $76\%$ of the generated adversarial examples maintain model-to-modeltransferability when the perturbation is applied to local image regions.Depending on the selected region, these localized adversarial examples requiresignificantly less $L_p$ norm distortion (for $p \in \{0, 2, \infty\}$)compared to their non-local counterparts. These localized attacks thereforehave the potential to undermine defenses that claim robustness under theaforementioned norms.

Regional Image Perturbation Reduces $L_p$ Norms of Adversarial Examples While Maintaining Model-to-model Transferability