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

Unsupervised Domain Adaptation (UDA) can effectively address domain gapissues in real-world image Super-Resolution (SR) by accessing both the sourceand target data. Considering privacy policies or transmission restrictions ofsource data in practical scenarios, we propose a SOurce-free Domain Adaptationframework for image SR (SODA-SR) to address this issue, i.e., adapt asource-trained model to a target domain with only unlabeled target data.SODA-SR leverages the source-trained model to generate refined pseudo-labelsfor teacher-student learning. To better utilize pseudo-labels, we propose anovel wavelet-based augmentation method, named Wavelet Augmentation Transformer(WAT), which can be flexibly incorporated with existing networks, to implicitlyproduce useful augmented data. WAT learns low-frequency information of varyinglevels across diverse samples, which is aggregated efficiently via deformableattention. Furthermore, an uncertainty-aware self-training mechanism isproposed to improve the accuracy of pseudo-labels, with inaccurate predictionsbeing rectified by uncertainty estimation. To acquire better SR results andavoid overfitting pseudo-labels, several regularization losses are proposed toconstrain target LR and SR images in the frequency domain. Experiments showthat without accessing source data, SODA-SR outperforms state-of-the-art UDAmethods in both synthetic$\rightarrow$real and real$\rightarrow$real adaptationsettings, and is not constrained by specific network architectures.

Uncertainty-Aware Source-Free Adaptive Image Super-Resolution with Wavelet Augmentation Transformer