Saved in:
Bibliographic Details
Main Authors: Cheng, Ching-Lam, Zhu, Bin, He, Shengfeng
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.24407
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912982193143808
author Cheng, Ching-Lam
Zhu, Bin
He, Shengfeng
author_facet Cheng, Ching-Lam
Zhu, Bin
He, Shengfeng
contents Generating realistic 3D hand motion from natural language is vital for VR, robotics, and human-computer interaction. Existing methods either focus on full-body motion, overlooking detailed hand gestures, or require explicit 3D object meshes, limiting generality. We propose TSHaMo, a model-agnostic teacher-student diffusion framework for text-driven hand motion generation. The student model learns to synthesize motions from text alone, while the teacher leverages auxiliary signals (e.g., MANO parameters) to provide structured guidance during training. A co-training strategy enables the student to benefit from the teacher's intermediate predictions while remaining text-only at inference. Evaluated using two diffusion backbones on GRAB and H2O, TSHaMo consistently improves motion quality and diversity. Ablations confirm its robustness and flexibility in using diverse auxiliary inputs without requiring 3D objects at test time.
format Preprint
id arxiv_https___arxiv_org_abs_2603_24407
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Teacher-Student Diffusion Model for Text-Driven 3D Hand Motion Generation
Cheng, Ching-Lam
Zhu, Bin
He, Shengfeng
Computer Vision and Pattern Recognition
Generating realistic 3D hand motion from natural language is vital for VR, robotics, and human-computer interaction. Existing methods either focus on full-body motion, overlooking detailed hand gestures, or require explicit 3D object meshes, limiting generality. We propose TSHaMo, a model-agnostic teacher-student diffusion framework for text-driven hand motion generation. The student model learns to synthesize motions from text alone, while the teacher leverages auxiliary signals (e.g., MANO parameters) to provide structured guidance during training. A co-training strategy enables the student to benefit from the teacher's intermediate predictions while remaining text-only at inference. Evaluated using two diffusion backbones on GRAB and H2O, TSHaMo consistently improves motion quality and diversity. Ablations confirm its robustness and flexibility in using diverse auxiliary inputs without requiring 3D objects at test time.
title Teacher-Student Diffusion Model for Text-Driven 3D Hand Motion Generation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2603.24407