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 Plus

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

Zendy Tools

Zendy Open

For monocular depth estimation, acquiring ground truths for real data is noteasy, and thus domain adaptation methods are commonly adopted using thesupervised synthetic data. However, this may still incur a large domain gap dueto the lack of supervision from the real data. In this paper, we develop adomain adaptation framework via generating reliable pseudo ground truths ofdepth from real data to provide direct supervisions. Specifically, we proposetwo mechanisms for pseudo-labeling: 1) 2D-based pseudo-labels via measuring theconsistency of depth predictions when images are with the same content butdifferent styles; 2) 3D-aware pseudo-labels via a point cloud completionnetwork that learns to complete the depth values in the 3D space, thusproviding more structural information in a scene to refine and generate morereliable pseudo-labels. In experiments, we show that our pseudo-labelingmethods improve depth estimation in various settings, including the usage ofstereo pairs during training. Furthermore, the proposed method performsfavorably against several state-of-the-art unsupervised domain adaptationapproaches in real-world datasets.

3D-PL: Domain Adaptive Depth Estimation with 3D-aware Pseudo-Labeling