Research Analyzer
← Back ICRA 2026

UniFucGrasp: Human-Hand-Inspired Unified Functional Grasp Annotation Strategy and Dataset for Diverse Dexterous Hands

Haoran Lin, Wenrui Chen, Xianchi Chen, Fan Yang, Qiang Diao, Wenxin Xie, Sijie Wu, Kailun Yang, Maojun Li, Yaonan Wang

PDF

AI summary

Key figure (auto-extracted from paper)
A unified human-to-robot mapping strategy enables efficient, stable functional grasp annotation across diverse fully- and under-actuated dexterous hands, validated on a new large-scale dataset and synthesis model.
functional grasping dexterous hands human-to-robot mapping grasp dataset biomimetic control robotic manipulation

Problem

Existing functional grasp datasets are restricted to expensive, fully-actuated robotic hands, creating high annotation costs and poor generalization across different hand architectures.

Approach

The authors introduce a biomimetic mapping framework that translates natural human hand motions into control commands for heterogeneous robotic hands using sparse matrix optimization and force-closure analysis.

Key results

  • Unified human-to-robot pose mapping bridging structural and actuation differences
  • UniFucGrasp dataset with 100K+ functional grasp annotations across 1,108 objects
  • Point cloud-conditioned gesture generation model for multi-hand grasping
  • Improved manipulation accuracy, stability, and cross-hand generalization in simulation and real-world tests

Why it matters

It provides a scalable, hardware-agnostic pipeline and dataset to accelerate adaptive dexterous manipulation research and reduce annotation bottlenecks.

Abstract

Dexterous grasp datasets are vital for embodied intelligence, but mostly emphasize grasp stability, ignoring func- tional grasps needed for tasks like opening bottle caps or holding cup handles. Most rely on bulky, costly, and hard- to-control high-DOF Shadow Hands. Inspired by the human hand’s underactuated mechanism, we establish UniFucGrasp, a universal functional grasp annotation strategy and dataset for multiple dexterous hand types. Based on biomimicry, it maps natural human motions to diverse hand structures and uses geometry-based force closure to ensure functional, stable, human- like grasps. This method supports low-cost, efficient collection of diverse, high-quality functional grasps. Finally, we establish the first multi-hand functional grasp dataset and provide a synthesis model to validate its effectiveness. Experiments on the UFG dataset, IsaacSim, and complex robotic tasks show that our method improves functional manipulation accuracy and grasp stability, demonstrates improved adaptability across multiple robotic hands, helping to alleviate annotation cost and generalization challenges in dexterous grasping. The project page is at https://haochen611.github.io/UFG.

Index terms

Multifingered Hands Dexterous Manipulation Deep Learning in Grasping and Manipulation

Related papers