UniT: Toward a Unified Physical Language for Human-to-Humanoid Policy Learning and World Modeling
cs.RO, cs.AI
Submitted: 2026-04-21
Updated: 2026-10-11
Comments: Project page: https://xpeng-robotics.github.io/unit/ ; Author-recommended BibTeX: https://xpeng-robotics.github.io/unit/UniT.bib . This version adds an ancillary citation file; the manuscript is unchanged
Project page: https://xpeng-robotics.github.io/unit
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- In-N-On: Scaling Egocentric Manipulation with in-the-wild and on-task Data
- DexWild: Dexterous Human Interactions for In-the-Wild Robot Policies
- EgoVLA: Learning Vision-Language-Action Models from Egocentric Human Videos
- Behavior Generation with Latent Actions
- FAST: Efficient Action Tokenization for Vision-Language-Action Models
- Latent Action Pretraining from Videos
- UniVLA: Learning to Act Anywhere with Task-centric Latent Actions
- METIS: Multi-Source Egocentric Training for Integrated Dexterous Vision-Language-Action Model
- R3M: A Universal Visual Representation for Robot Manipulation
- VIP: Towards Universal Visual Reward and Representation via Value-Implicit Pre-Training
- Humanoid Policy ~ Human Policy
- MotionTrans: Human VR Data Enable Motion-Level Learning for Robotic Manipulation Policies
- Emergence of Human to Robot Transfer in Vision-Language-Action Models
- villa-X: Enhancing Latent Action Modeling in Vision-Language-Action Models
- XR-1: Towards Versatile Vision-Language-Action Models via Learning Unified Vision-Motion Representations
- Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success
- $\pi_0$: A Vision-Language-Action Flow Model for General Robot Control
- Octo: An Open-Source Generalist Robot Policy
- Ctrl-World: A Controllable Generative World Model for Robot Manipulation
- DINOv2: Learning Robust Visual Features without Supervision
Related papers
- FMT x: An Efficient and Asymptotically Optimal Extension of the Fast Marching Tree for Dynamic Replanning
- MPCFormer: A physics-informed data-driven approach for explainable socially-aware autonomous driving
- RoboLab: A High-Fidelity Simulation Benchmark for Analysis of Task Generalist Policies
- ARC: A Reasoning Recipe for Robot Foundation Models
- HRDexDB: A 4D Dexterous Grasping Dataset Across Human and Multiple Robot Embodiments
- APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies