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

We study adversarial examples in a black-box setting where the adversary onlyhas API access to the target model and each query is expensive. Prior work onblack-box adversarial examples follows one of two main strategies: (1) transferattacks use white-box attacks on local models to find candidate adversarialexamples that transfer to the target model, and (2) optimization-based attacksuse queries to the target model and apply optimization techniques to search foradversarial examples. We propose hybrid attacks that combine both strategies,using candidate adversarial examples from local models as starting points foroptimization-based attacks and using labels learned in optimization-basedattacks to tune local models for finding transfer candidates. We empiricallydemonstrate on the MNIST, CIFAR10, and ImageNet datasets that our hybrid attackstrategy reduces cost and improves success rates. We also introduce a seedprioritization strategy which enables attackers to focus their resources on themost promising seeds. Combining hybrid attacks with our seed prioritizationstrategy enables batch attacks that can reliably find adversarial examples withonly a handful of queries.

Hybrid Batch Attacks: Finding Black-box Adversarial Examples with Limited Queries