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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Many optimization methods for generating black-box adversarial examples havebeen proposed, but the aspect of initializing said optimizers has not beenconsidered in much detail. We show that the choice of starting points is indeedcrucial, and that the performance of state-of-the-art attacks depends on it.First, we discuss desirable properties of starting points for attacking imageclassifiers, and how they can be chosen to increase query efficiency. Notably,we find that simply copying small patches from other images is a validstrategy. We then present an evaluation on ImageNet that clearly demonstratesthe effectiveness of this method: Our initialization scheme reduces the numberof queries required for a state-of-the-art Boundary Attack by 81%,significantly outperforming previous results reported for targeted black-boxadversarial examples.

Copy and Paste: A Simple But Effective Initialization Method for Black-Box Adversarial Attacks