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

6D object pose estimation aims to infer the relative pose between the objectand the camera using a single image or multiple images. Most works have focusedon predicting the object pose without associated uncertainty under occlusionand structural ambiguity (symmetricity). However, these works demand priorinformation about shape attributes, and this condition is hardly satisfied inreality; even asymmetric objects may be symmetric under the viewpoint change.In addition, acquiring and fusing diverse sensor data is challenging whenextending them to robotics applications. Tackling these limitations, we presentan ambiguity-aware 6D object pose estimation network, PrimA6D++, as a genericuncertainty prediction method. The major challenges in pose estimation, such asocclusion and symmetry, can be handled in a generic manner based on themeasured ambiguity of the prediction. Specifically, we devise a network toreconstruct the three rotation axis primitive images of a target object andpredict the underlying uncertainty along each primitive axis. Leveraging theestimated uncertainty, we then optimize multi-object poses using visualmeasurements and camera poses by treating it as an object SLAM problem. Theproposed method shows a significant performance improvement in T-LESS andYCB-Video datasets. We further demonstrate real-time scene recognitioncapability for visually-assisted robot manipulation. Our code and supplementarymaterials are available at https://github.com/rpmsnu/PrimA6D.

Ambiguity-Aware Multi-Object Pose Optimization for Visually-Assisted Robot Manipulation