Pixel-UMI: Learning Environment-Aware Whole-Body Manipulation from Sensorized Human Demonstrations
cs.RO, cs.AI, cs.CV, cs.LG, cs.SY, eess.SY
Submitted: 2026-11-02
Updated: 2026-11-02
Comments: Video results are available at https://pixel-robot.com/pixel-umi
Code: https://github.com/kevinzakka/mink
Project page: https://mobile-aloha.github.io
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- HoMMI: Learning Whole-Body Mobile Manipulation from Human Demonstrations
- ActiveUMI: Robotic Manipulation with Active Perception from Robot-Free Human Demonstrations
- Geometric Fabrics for the Acceleration-based Design of Robotic Motion
- ForceMimic: Force-Centric Imitation Learning with Force-Motion Capture System for Contact-Rich Manipulation
- In-the-Wild Compliant Manipulation with UMI-FT
- HG-DAgger: Interactive Imitation Learning with Human Experts
- RaC: Robot Learning for Long-Horizon Tasks by Scaling Recovery and Correction
- Training-Time Action Conditioning for Efficient Real-Time Chunking
- DINOv3
- Flow Matching for Generative Modeling
- Grasping in the Wild:Learning 6DoF Closed-Loop Grasping from Low-Cost Demonstrations
- On Bringing Robots Home
- YUBI: Yielding Universal Bidigital Interface for Bimanual Dexterous Manipulation at Scale
- AirExo-2: Scaling up Generalizable Robotic Imitation Learning with Low-Cost Exoskeletons
- DexUMI: Using Human Hand as the Universal Manipulation Interface for Dexterous Manipulation
- EgoMimic: Scaling Imitation Learning via Egocentric Video
- EgoMI: Learning Active Vision and Whole-Body Manipulation from Egocentric Human Demonstrations
- HuMI: Humanoid Whole-Body Manipulation from Humanoid-Free Demonstrations
- HALOMI: Learning Humanoid Loco-Manipulation with Active Perception from Human Demonstrations
- MoMaGen: Generating Demonstrations under Soft and Hard Constraints for Multi-Step Bimanual Mobile Manipulation
Related papers
- Rethinking Actions: Continuity-Aware Priors for World-Model Planning
- MOSAIC: Modular Obstacle and Spatial-Aware Framework for Manipulation In Cluttered Scenes
- SAGAS: Semantic-Aware Graph-Assisted Stitching for Offline Temporal Logic Planning
- DexSeed: Expanding Human Hand-Object Interactions into Robot Demonstrations for Dexterous Policy Learning
- LaST0.5: A Generalist World Action Model with Interaction-Centric Latent Reasoning
- CIALLO: Learning Shared Intent Representations for Cross-Embodiment Manipulation