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 network robustness has become a central topic in machine learning inrecent years. Most training algorithms that improve the model's robustness toadversarial and common corruptions also introduce a large computationaloverhead, requiring as many as ten times the number of forward and backwardpasses in order to converge. To combat this inefficiency, we proposeBulletTrain $-$ a boundary example mining technique to drastically reduce thecomputational cost of robust training. Our key observation is that only a smallfraction of examples are beneficial for improving robustness. BulletTraindynamically predicts these important examples and optimizes robust trainingalgorithms to focus on the important examples. We apply our technique toseveral existing robust training algorithms and achieve a 2.1$\times$ speed-upfor TRADES and MART on CIFAR-10 and a 1.7$\times$ speed-up for AugMix onCIFAR-10-C and CIFAR-100-C without any reduction in clean and robust accuracy.

BulletTrain: Accelerating Robust Neural Network Training via Boundary Example Mining