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

Vision Transformer (ViT) is becoming more popular in image processing.Specifically, we investigate the effectiveness of test-time adaptation (TTA) onViT, a technique that has emerged to correct its prediction during test-time byitself. First, we benchmark various test-time adaptation approaches on ViT-B16and ViT-L16. It is shown that the TTA is effective on ViT and theprior-convention (sensibly selecting modulation parameters) is not necessarywhen using proper loss function. Based on the observation, we propose a newtest-time adaptation method called class-conditional feature alignment (CFA),which minimizes both the class-conditional distribution differences and thewhole distribution differences of the hidden representation between the sourceand target in an online manner. Experiments of image classification tasks oncommon corruption (CIFAR-10-C, CIFAR-100-C, and ImageNet-C) and domainadaptation (digits datasets and ImageNet-Sketch) show that CFA stablyoutperforms the existing baselines on various datasets. We also verify that CFAis model agnostic by experimenting on ResNet, MLP-Mixer, and several ViTvariants (ViT-AugReg, DeiT, and BeiT). Using BeiT backbone, CFA achieves 19.8%top-1 error rate on ImageNet-C, outperforming the existing test-time adaptationbaseline 44.0%. This is a state-of-the-art result among TTA methods that do notneed to alter training phase.

Robustifying Vision Transformer without Retraining from Scratch by Test-Time Class-Conditional Feature Alignment