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

Deep neural networks achieve high prediction accuracy when the train and testdistributions coincide. In practice though, various types of corruptions occurwhich deviate from this setup and cause severe performance degradations. Fewmethods have been proposed to address generalization in the presence ofunforeseen domain shifts. In particular, digital noise corruptions arisecommonly in practice during the image acquisition stage and present asignificant challenge for current methods. In this paper, we propose a diverseGaussian noise consistency regularization method for improving robustness ofimage classifiers under a variety of corruptions while still maintaining highclean accuracy. We derive bounds to motivate and understand the behavior of ourGaussian noise consistency regularization using a local loss landscapeanalysis. Our approach improves robustness against unforeseen noise corruptionsby 4.2-18.4% over adversarial training and other strong diverse dataaugmentation baselines across several benchmarks. Furthermore, it improvesrobustness and uncertainty calibration by 3.7% and 5.5%, respectively, againstall common corruptions (weather, digital, blur, noise) when combined withstate-of-the-art diverse data augmentations.

Diverse Gaussian Noise Consistency Regularization for Robustness and Uncertainty Calibration