Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
September 29, 2025Open Access

OpenHOI: Open-World Hand-Object Interaction Synthesis with Multimodal Large Language Model

View Full Paper
Ask AI
Bookmark
Share

Authors

ZZZhenhao ZhangSYShi YeLYLingxiao Yang

Discussion

Loading...

Member takes

Overview

OpenHOI generates realistic hand-object interactions in 3D environments, suggesting new solutions for robotics and AR/VR.

Key Points

  • OpenHOI synthesizes long-horizon manipulation sequences, improving generalization to novel objects and tasks.
  • The framework outperforms existing methods by leveraging a multimodal language model for affordance grounding.
  • An affordance-driven diffusion model is used to optimize physical interactions and minimize errors.
  • Evaluations show significant improvements in handling complex language instructions and multi-stage tasks.

Cite This Study

Zhang et al. (2025) studied this question.

synapsesocial.com/papers/68da58d8c1728099cfd11269https://doi.org/10.48550/arxiv.2505.18947
View Full Paper
Ask AI
Bookmark
Share