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

Scene representation has been a crucial design choice in robotic manipulationsystems. An ideal representation should be 3D, dynamic, and semantic to meetthe demands of diverse manipulation tasks. However, previous works often lackall three properties simultaneously. In this work, we introduce D$^3$Fields -dynamic 3D descriptor fields. These fields capture the dynamics of theunderlying 3D environment and encode both semantic features and instance masks.Specifically, we project arbitrary 3D points in the workspace onto multi-view2D visual observations and interpolate features derived from foundationalmodels. The resulting fused descriptor fields allow for flexible goalspecifications using 2D images with varied contexts, styles, and instances. Toevaluate the effectiveness of these descriptor fields, we apply ourrepresentation to a wide range of robotic manipulation tasks in a zero-shotmanner. Through extensive evaluation in both real-world scenarios andsimulations, we demonstrate that D$^3$Fields are both generalizable andeffective for zero-shot robotic manipulation tasks. In quantitative comparisonswith state-of-the-art dense descriptors, such as Dense Object Nets and DINO,D$^3$Fields exhibit significantly better generalization abilities andmanipulation accuracy.

D$^3$Fields: Dynamic 3D Descriptor Fields for Zero-Shot Generalizable Robotic Manipulation