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

Robust machine learning is currently one of the most prominent topics whichcould potentially help shaping a future of advanced AI platforms that not onlyperform well in average cases but also in worst cases or adverse situations.Despite the long-term vision, however, existing studies on black-boxadversarial attacks are still restricted to very specific settings of threatmodels (e.g., single distortion metric and restrictive assumption on targetmodel's feedback to queries) and/or suffer from prohibitively high querycomplexity. To push for further advances in this field, we introduce a generalframework based on an operator splitting method, the alternating directionmethod of multipliers (ADMM) to devise efficient, robust black-box attacks thatwork with various distortion metrics and feedback settings without incurringhigh query complexity. Due to the black-box nature of the threat model, theproposed ADMM solution framework is integrated with zeroth-order (ZO)optimization and Bayesian optimization (BO), and thus is applicable to thegradient-free regime. This results in two new black-box adversarial attackgeneration methods, ZO-ADMM and BO-ADMM. Our empirical evaluations on imageclassification datasets show that our proposed approaches have much lowerfunction query complexities compared to state-of-the-art attack methods, butachieve very competitive attack success rates.

On the Design of Black-box Adversarial Examples by Leveraging Gradient-free Optimization and Operator Splitting Method