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/

Zendy Open

Presents the pdf analysis as premium feature

PDF Analysis

Insights

Presents the summarisation as premium feature

Summarisation

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Generalizing knowledge to unseen domains, where data and labels areunavailable, is crucial for machine learning models. We tackle the domaingeneralization problem to learn from multiple source domains and generalize toa target domain with unknown statistics. The crucial idea is to extract theunderlying invariant features across all the domains. Previous domaingeneralization approaches mainly focused on learning invariant features andstacking the learned features from each source domain to generalize to a newtarget domain while ignoring the label information, which will lead toindistinguishable features with an ambiguous classification boundary. For this,one possible solution is to constrain the label-similarity when extracting theinvariant features and to take advantage of the label similarities forclass-specific cohesion and separation of features across domains. Therefore weadopt optimal transport with Wasserstein distance, which could constrain theclass label similarity, for adversarial training and also further deploy ametric learning objective to leverage the label information for achievingdistinguishable classification boundary. Empirical results show that ourproposed method could outperform most of the baselines. Furthermore, ablationstudies also demonstrate the effectiveness of each component of our method.

Domain Generalization via Optimal Transport with Metric Similarity Learning