DexFlow: A Unified Approach for Dexterous Hand Pose Retargeting and Interaction
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866916717126483968 |
|---|---|
| author | Lin, Xiaoyi Yao, Kunpeng Xu, Lixin Wang, Xueqiang Li, Xuetao Wang, Yuchen Li, Miao |
| author_facet | Lin, Xiaoyi Yao, Kunpeng Xu, Lixin Wang, Xueqiang Li, Xuetao Wang, Yuchen Li, Miao |
| contents | Despite advances in hand-object interaction modeling, generating realistic dexterous manipulation data for robotic hands remains a challenge. Retargeting methods often suffer from low accuracy and fail to account for hand-object interactions, leading to artifacts like interpenetration. Generative methods, lacking human hand priors, produce limited and unnatural poses. We propose a data transformation pipeline that combines human hand and object data from multiple sources for high-precision retargeting. Our approach uses a differential loss constraint to ensure temporal consistency and generates contact maps to refine hand-object interactions. Experiments show our method significantly improves pose accuracy, naturalness, and diversity, providing a robust solution for hand-object interaction modeling. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_01083 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DexFlow: A Unified Approach for Dexterous Hand Pose Retargeting and Interaction Lin, Xiaoyi Yao, Kunpeng Xu, Lixin Wang, Xueqiang Li, Xuetao Wang, Yuchen Li, Miao Robotics Despite advances in hand-object interaction modeling, generating realistic dexterous manipulation data for robotic hands remains a challenge. Retargeting methods often suffer from low accuracy and fail to account for hand-object interactions, leading to artifacts like interpenetration. Generative methods, lacking human hand priors, produce limited and unnatural poses. We propose a data transformation pipeline that combines human hand and object data from multiple sources for high-precision retargeting. Our approach uses a differential loss constraint to ensure temporal consistency and generates contact maps to refine hand-object interactions. Experiments show our method significantly improves pose accuracy, naturalness, and diversity, providing a robust solution for hand-object interaction modeling. |
| title | DexFlow: A Unified Approach for Dexterous Hand Pose Retargeting and Interaction |
| topic | Robotics |
| url | https://arxiv.org/abs/2505.01083 |