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

The majority of existing Unsupervised Domain Adaptation (UDA) methodspresumes source and target domain data to be simultaneously available duringtraining. Such an assumption may not hold in practice, as source data is ofteninaccessible (e.g., due to privacy reasons). On the contrary, a pre-trainedsource model is always considered to be available, even though performingpoorly on target due to the well-known domain shift problem. This translatesinto a significant amount of misclassifications, which can be interpreted asstructured noise affecting the inferred target pseudo-labels. In this work, wecast UDA as a pseudo-label refinery problem in the challenging source-freescenario. We propose a unified method to tackle adaptive noise filtering andpseudo-label refinement. A novel Negative Ensemble Learning technique isdevised to specifically address noise in pseudo-labels, by enhancing diversityin ensemble members with different stochastic (i) input augmentation and (ii)feedback. In particular, the latter is achieved by leveraging the novel conceptof Disjoint Residual Labels, which allow diverse information to be fed to thedifferent members. A single target model is eventually trained with the refinedpseudo-labels, which leads to a robust performance on the target domain.Extensive experiments show that the proposed method, named AdaptivePseudo-Label Refinement, achieves state-of-the-art performance on major UDAbenchmarks, such as Digit5, PACS, Visda-C, and DomainNet, without using sourcedata at all.

Adaptive Pseudo-Label Refinement by Negative Ensemble Learning for Source-Free Unsupervised Domain Adaptation