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

We propose a physics-based method for synthesizing dexterous hand-objectinteractions in a full-body setting. While recent advancements have addressedspecific facets of human-object interactions, a comprehensive physics-basedapproach remains a challenge. Existing methods often focus on isolated segmentsof the interaction process and rely on data-driven techniques that may resultin artifacts. In contrast, our proposed method embraces reinforcement learning(RL) and physics simulation to mitigate the limitations of data-drivenapproaches. Through a hierarchical framework, we first learn skill priors forboth body and hand movements in a decoupled setting. The generic skill priorslearn to decode a latent skill embedding into the motion of the underlyingpart. A high-level policy then controls hand-object interactions in thesepretrained latent spaces, guided by task objectives of grasping and 3D targettrajectory following. It is trained using a novel reward function that combinesan adversarial style term with a task reward, encouraging natural motions whilefulfilling the task incentives. Our method successfully accomplishes thecomplete interaction task, from approaching an object to grasping andsubsequent manipulation. We compare our approach against kinematics-basedbaselines and show that it leads to more physically plausible motions.

Physically Plausible Full-Body Hand-Object Interaction Synthesis