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

The rapid development of inexpensive commodity depth sensors has madekeypoint detection and matching in the depth image modality an importantproblem in computer vision. Despite great improvements in recent RGB localfeature learning methods, adapting them directly in the depth modality leads tounsatisfactory performance. Most of these methods do not explicitly reasonbeyond the visible pixels in the images. To address the limitations of thesemethods, we propose a framework ViewSynth, to jointly learn: (1) viewpointinvariant keypoint-descriptor from depth images using a proposed ContrastiveMatching Loss, and (2) view synthesis of depth images from different viewpointsusing the proposed View Synthesis Module and View Synthesis Loss. By learningview synthesis, we explicitly encourage the feature extractor to encodeinformation about not only the visible, but also the occluded parts of thescene. We demonstrate that in the depth modality, ViewSynth outperforms thestate-of-the-art depth and RGB local feature extraction techniques in the 3Dkeypoint matching and camera localization tasks on the RGB-D datasets 7-Scenes,TUM RGBD and CoRBS in most scenarios. We also show the generalizability ofViewSynth in 3D keypoint matching across different datasets.

ViewSynth: Learning Local Features from Depth using View Synthesis