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

Neural networks have been proven to be both highly effective within computervision, and highly vulnerable to adversarial attacks. Consequently, as the useof neural networks increases due to their unrivaled performance, so too doesthe threat posed by adversarial attacks. In this work, we build towardsaddressing the challenge of adversarial robustness by exploring therelationship between the mini-batch size used during adversarial samplegeneration and the strength of the adversarial samples produced. We demonstratethat an increase in mini-batch size results in a decrease in the efficacy ofthe samples produced, and we draw connections between these observations andthe phenomenon of vanishing gradients. Next, we formulate loss functions suchthat adversarial sample strength is not degraded by mini-batch size. Ourfindings highlight a potential risk for underestimating the true (practical)strength of adversarial attacks, and a risk of overestimating a model'srobustness. We share our codes to let others replicate our experiments and tofacilitate further exploration of the connections between batch size andadversarial sample strength.

Generating Adversarial Samples in Mini-Batches May Be Detrimental To Adversarial Robustness