Research map / Vision-language and multimodal / Robot action and navigation
UniDex: A Robot Foundation Suite for Universal Dexterous Hand Control from Egocentric Human Videos
Gu Zhang, Qicheng Xu, Haozhe Zhang, Jianhan Ma, Long He, Yiming Bao, Zeyu Ping, Zhecheng Yuan, Chenhao Lu, Chengbo Yuan, Tianhai Liang, Xiaoyu Tian and 7 more
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| pretrain on UniDex-Dataset (vs no pretraining) (ablation) | UniDex-VLA · 5 real-world tool-use tasks | average task progress (%): 32.5 -> 81.0 | Improved |
| pretrain on UniDex-Dataset (vs no pretraining) (ablation) | UniDex-VLA · Zero-shot hand transfer to Wuji | average task progress (%): 0% -> 40% | Improved |
| pretrain on UniDex-Dataset (vs no pretraining) (ablation) | UniDex-VLA · Zero-shot hand transfer to Oymotion | average task progress (%): 5% -> 60% | Improved |
| 3D pointcloud VLA with FAAS vs pi0 baseline | UniDex-VLA · 5 real-world tool-use tasks | average task progress (%): 38.0 -> 81.0 | Improved |
| add UniDex-VLA vs DP3 diffusion baseline | UniDex-VLA · 5 real-world tool-use tasks | average task progress (%): 35.0 -> 81.0 | Improved |
| add DemoGen pointcloud augmentation for OOD placements | UniDex-VLA · Make Coffee spatial generalization | average task progress | Improved |
| co-train with transformed human demos (fixed robot demos) (ablation) | UniDex-VLA · Make Coffee | average task progress | Improved |
| human demos alone, without robot demos (ablation) | UniDex-VLA · Make Coffee | success: 0 | Hurt |
Closest papers by meaning
- HandEdit: A Unified Benchmark for Egocentric Human-to-Robot Dexterous Hand Image Editing (NeurIPS 2026 · similarity 0.92)
- Dexterous Manipulation Transfer via Progressive Kinematic-Dynamic Alignment (AAAI 2026 · similarity 0.92)
- ManipTrans: Efficient Dexterous Bimanual Manipulation Transfer via Residual Learning (CVPR 2025 · similarity 0.89)
- Scaffolding Dexterous Manipulation with Vision-Language Models (NeurIPS 2025 · similarity 0.89)
- Cross-Embodiment Dexterous Grasping with Reinforcement Learning (ICLR 2025 · similarity 0.89)
- StableHand: Quality-Aware Flow Matching for World-Space Dual-Hand Motion Estimation from Egocentric Video (NeurIPS 2026 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
