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

Adversarial learning baselines for domain adaptation (DA) approaches in thecontext of semantic segmentation are under explored in semi-supervisedframework. These baselines involve solely the available labeled target samplesin the supervision loss. In this work, we propose to enhance their usefulnesson both semantic segmentation and the single domain classifier neural networks.We design new training objective losses for cases when labeled target databehave as source samples or as real target samples. The underlying rationale isthat considering the set of labeled target samples as part of source domainhelps reducing the domain discrepancy and, hence, improves the contribution ofthe adversarial loss. To support our approach, we consider a complementarymethod that mixes source and labeled target data, then applies the sameadaptation process. We further propose an unsupervised selection procedureusing entropy to optimize the choice of labeled target samples for adaptation.We illustrate our findings through extensive experiments on the benchmarksGTA5, SYNTHIA, and Cityscapes. The empirical evaluation highlights competitiveperformance of our proposed approach.

Adversarial Semi-Supervised Domain Adaptation for Semantic Segmentation: A New Role for Labeled Target Samples