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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Recent study of adversarial attacks has revealed the vulnerability of moderndeep learning models. That is, subtly crafted perturbations of the input canmake a trained network with high accuracy produce arbitrary incorrectpredictions, while maintain imperceptible to human vision system. In thispaper, we introduce Block Switching (BS), a defense strategy againstadversarial attacks based on stochasticity. BS replaces a block of model layerswith multiple parallel channels, and the active channel is randomly assigned inthe run time hence unpredictable to the adversary. We show empirically that BSleads to a more dispersed input gradient distribution and superior defenseeffectiveness compared with other stochastic defenses such as stochasticactivation pruning (SAP). Compared to other defenses, BS is also characterizedby the following features: (i) BS causes less test accuracy drop; (ii) BS isattack-independent and (iii) BS is compatible with other defenses and can beused jointly with others.

Block Switching: A Stochastic Approach for Deep Learning Security