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

Contemporary monocular 6D pose estimation methods can only cope with ahandful of object instances. This naturally hampers possible applications as,for instance, robots seamlessly integrated in everyday processes necessarilyrequire the ability to work with hundreds of different objects. To tackle thisproblem of immanent practical relevance, we propose a novel method forclass-level monocular 6D pose estimation, coupled with metric shape retrieval.Unfortunately, acquiring adequate annotations is very time-consuming and laborintensive. This is especially true for class-level 6D pose estimation, as oneis required to create a highly detailed reconstruction for all objects and thenannotate each object and scene using these models. To overcome thisshortcoming, we additionally propose the idea of synthetic-to-real domaintransfer for class-level 6D poses by means of self-supervised learning, whichremoves the burden of collecting numerous manual annotations. In essence, aftertraining our proposed method fully supervised with synthetic data, we leveragerecent advances in differentiable rendering to self-supervise the model withunannotated real RGB-D data to improve latter inference. We experimentallydemonstrate that we can retrieve precise 6D poses and metric shapes from asingle RGB image.

CPS++: Improving Class-level 6D Pose and Shape Estimation From Monocular Images With Self-Supervised Learning