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

Robotic fabric manipulation is challenging due to the infinite dimensionalconfiguration space, self-occlusion, and complex dynamics of fabrics. There hasbeen significant prior work on learning policies for specific deformablemanipulation tasks, but comparatively less focus on algorithms which canefficiently learn many different tasks. In this paper, we learn visualcorrespondences for deformable fabrics across different configurations insimulation and show that this representation can be used to design policies fora variety of tasks. Given a single demonstration of a new task from an initialfabric configuration, the learned correspondences can be used to computegeometrically equivalent actions in a new fabric configuration. This makes itpossible to robustly imitate a broad set of multi-step fabric smoothing andfolding tasks on multiple physical robotic systems. The resulting policiesachieve 80.3% average task success rate across 10 fabric manipulation tasks ontwo different robotic systems, the da Vinci surgical robot and the ABB YuMi.Results also suggest robustness to fabrics of various colors, sizes, andshapes. See https://tinyurl.com/fabric-descriptors for supplementary materialand videos.

Learning Dense Visual Correspondences in Simulation to Smooth and Fold Real Fabrics