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

Learning deep neural networks that are generalizable across different domainsremains a challenge due to the problem of domain shift. Unsupervised domainadaptation is a promising avenue which transfers knowledge from a source domainto a target domain without using any labels in the target domain. Contemporarytechniques focus on extracting domain-invariant features using domainadversarial training. However, these techniques neglect to learn discriminativeclass boundaries in the latent representation space on a target domain andyield limited adaptation performance. To address this problem, we proposedistance metric guided feature alignment (MetFA) to extract discriminative aswell as domain-invariant features on both source and target domains. Theproposed MetFA method explicitly and directly learns the latent representationwithout using domain adversarial training. Our model integrates classdistribution alignment to transfer semantic knowledge from a source domain to atarget domain. We evaluate the proposed method on fetal ultrasound datasets forcross-device image classification. Experimental results demonstrate that theproposed method outperforms the state-of-the-art and enables modelgeneralization.

Unsupervised Cross-domain Image Classification by Distance Metric Guided Feature Alignment