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

Manipulation planning is the task of computing robot trajectories that move aset of objects to their target configuration while satisfying physicallyfeasibility. In contrast to existing works that assume known object templates,we are interested in manipulation planning for a category of objects withpotentially unknown instances and large intra-category shape variation. Toachieve it, we need an object representation with which the manipulationplanner can reason about both the physical feasibility and desired objectconfiguration, while being generalizable to novel instances. The widely-usedpose representation is not suitable, as representing an object with aparameterized transformation from a fixed template cannot capture largeintra-category shape variation. Hence, we propose a new hybrid objectrepresentation consisting of semantic keypoint and dense geometry (a pointcloud or mesh) as the interface between the perception module and motionplanner. Leveraging advances in learning-based keypoint detection and shapecompletion, both dense geometry and keypoints can be perceived from raw sensorinput. Using the proposed hybrid object representation, we formulate themanipulation task as a motion planning problem which encodes both the objecttarget configuration and physical feasibility for a category of objects. Inthis way, many existing manipulation planners can be generalized to categoriesof objects, and the resulting perception-to-action manipulation pipeline isrobust to large intra-category shape variation. Extensive hardware experimentsdemonstrate our pipeline can produce robot trajectories that accomplish taskswith never-before-seen objects.

kPAM-SC: Generalizable Manipulation Planning using KeyPoint Affordance and Shape Completion