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

We address the task of estimating 6D camera poses from sparse-view image sets(2-8 images). This task is a vital pre-processing stage for nearly allcontemporary (neural) reconstruction algorithms but remains challenging givensparse views, especially for objects with visual symmetries and texture-lesssurfaces. We build on the recent RelPose framework which learns a network thatinfers distributions over relative rotations over image pairs. We extend thisapproach in two key ways; first, we use attentional transformer layers toprocess multiple images jointly, since additional views of an object mayresolve ambiguous symmetries in any given image pair (such as the handle of amug that becomes visible in a third view). Second, we augment this network toalso report camera translations by defining an appropriate coordinate systemthat decouples the ambiguity in rotation estimation from translationprediction. Our final system results in large improvements in 6D poseprediction over prior art on both seen and unseen object categories and alsoenables pose estimation and 3D reconstruction for in-the-wild objects.

RelPose++: Recovering 6D Poses from Sparse-view Observations