Function beyond Form: Functional Correspondence for Cross-Embodiment Dexterous Grasp Generation

2026-09-30

AI Summary

This study addresses the challenge of generalizing dexterous grasping across heterogeneous robotic hands caused by geometric and topological discrepancies. To this end, it proposes a functional alignment mechanism that maps physical links to shared functional components within a canonical coordinate system, integrating diffusion models with kinematic transformations to generate executable joint configurations. The core innovation lies in enabling knowledge transfer without requiring target-hand data or fine-tuning, necessitating only lightweight annotations. Experimental evaluations demonstrate that the proposed method achieves grasping success rates of 74.02% and 76.00% on unseen hands in simulation and real-world settings, respectively, validating its effectiveness for cross-embodiment generalization.

Abstract

Cross-embodiment dexterous grasp generation remains challenging because robotic hands differ substantially in geometry, topology, and kinematics. Existing approaches often lack explicit correspondences between structurally different hand regions that play similar functional roles in a grasp, a concept we refer to as functional correspondence. Consequently, their models tend to learn hand-specific interaction patterns rather than transferable grasp knowledge, limiting generalization to unseen hands. To address this limitation, we introduce FunCo-Grasp, which establishes functional correspondences across heterogeneous hand embodiments. Specifically, Functional Part Alignment aligns each hand to a canonical functional schema by mapping physical links to shared functional parts according to their grasping roles, while Canonical Frame Alignment expresses these parts in canonical local frames. These two alignments provide a consistent representation for inter-part and hand-object interactions, allowing the model to learn transferable grasp knowledge across hands. Conditioned on the aligned hand representation and object geometry, a diffusion model generates the target spatial arrangement of the functional parts, which are then converted into an executable joint configuration. Adapting FunCo-Grasp to an unseen hand requires only its geometric and kinematic models and a one-time lightweight functional annotation, without target-hand grasp data, fine-tuning, or learned retargeting. In simulation on held-out objects from the filtered CMapDataset, we achieves average success rates of 92.40% on three seen hands and 74.02% on four unseen hands. In real-world experiments, the same model achieves an overall success rate of 76.00% on two unseen hands without additional training or fine-tuning. These results demonstrate the effectiveness of FunCo-Grasp in transferring grasp knowledge to unseen hands.

Problem

• cross-embodiment
• dexterous grasp generation
• functional correspondence
• generalization
• transferable grasp knowledge

Innovation

• Cross-embodiment Dexterous Grasp
• Functional Correspondence
• Diffusion Model
• Zero-shot Generalization
• Canonical Alignment