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

In this work, we aim to learn dexterous manipulation of deformable objectsusing multi-fingered hands. Reinforcement learning approaches for dexterousrigid object manipulation would struggle in this setting due to the complexityof physics interaction with deformable objects. At the same time, previoustrajectory optimization approaches with differentiable physics for deformablemanipulation would suffer from local optima caused by the explosion of contactmodes from hand-object interactions. To address these challenges, we proposeDexDeform, a principled framework that abstracts dexterous manipulation skillsfrom human demonstration and refines the learned skills with differentiablephysics. Concretely, we first collect a small set of human demonstrations usingteleoperation. And we then train a skill model using demonstrations forplanning over action abstractions in imagination. To explore the goal space, wefurther apply augmentations to the existing deformable shapes in demonstrationsand use a gradient optimizer to refine the actions planned by the skill model.Finally, we adopt the refined trajectories as new demonstrations for finetuningthe skill model. To evaluate the effectiveness of our approach, we introduce asuite of six challenging dexterous deformable object manipulation tasks.Compared with baselines, DexDeform is able to better explore and generalizeacross novel goals unseen in the initial human demonstrations.

DexDeform: Dexterous Deformable Object Manipulation with Human Demonstrations and Differentiable Physics