Ropedia W0: Scaling Loco-Manipulation World Models across Embodiments with Human Experience
Ropedia W0: Scaling Loco-Manipulation World Models across Embodiments with Human Experience
This research presents Ropedia W0, the first egocentric loco-manipulation world foundation model spanning humans and robots with diverse embodiments. The model is trained on 10k hours of human egocentric data and 3,000 hours of robot data with unified multimodal annotations. It achieves state-of-the-art performance across five cross-embodiment benchmarks and supports real-time VR interaction after distillation.
The paper introduces Ropedia W0, a world model for loco-manipulation across human and robot embodiments, leveraging 10k hours of egocentric human data and 3k hours of robot data with unified annotations. It demonstrates cross-embodiment performance on multiple benchmarks and supports real-time VR interaction after distillation.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
Reported numbers
egocentric human data hours
10,000 hours
View original evidence
ents. We curate Xp-Data, comprising 10k hours of egocentric human data and 3k hours of robot data with unified multimodal annotations and systematic quality control. A unified kinematic representation encodes locomotion through camera poses and manipulation through projected skeletons, with robot skeletons combining embodiment-specific forward kinematics with extensive manual keypoint annotation and correction. We al
Open source S3robot data hours
3,000 hours
View original evidence
ents. We curate Xp-Data, comprising 10k hours of egocentric human data and 3k hours of robot data with unified multimodal annotations and systematic quality control. A unified kinematic representation encodes locomotion through camera poses and manipulation through projected skeletons, with robot skeletons combining embodiment-specific forward kinematics with extensive manual keypoint annotation and correction. We al
Open source S3cross-embodiment benchmarks
5
View original evidence
so introduce Xp-Teleop, a long-horizon loco-manipulation benchmark based on high-quality dexterous humanoid teleoperation data. Evaluations demonstrate state-of-the-art performance across five cross-embodiment benchmarks (DreamDojo, $\Psi_0$, $\omega$-HOME, Xp-Teleop, and AgiBot) and on held-out human egocentric video, from zero-shot inference to few-shot and full-data post-training. Scaling human pre-training data i
Open source S4
- paper: not_reported
- code: not_reported
- weights: not_reported
- dataset: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
ents. We curate Xp-Data, comprising 10k hours of egocentric human data and 3k hours of robot data with unified multimodal annotations and systematic quality control. A unified kinematic representation encodes locomotion through camera poses and manipulation through projected skeletons, with robot skeletons combining embodiment-specific forward kinematics with extensive manual keypoint annotation and correction. We al
Open source S3
so introduce Xp-Teleop, a long-horizon loco-manipulation benchmark based on high-quality dexterous humanoid teleoperation data. Evaluations demonstrate state-of-the-art performance across five cross-embodiment benchmarks (DreamDojo, $\Psi_0$, $\omega$-HOME, Xp-Teleop, and AgiBot) and on held-out human egocentric video, from zero-shot inference to few-shot and full-data post-training. Scaling human pre-training data i
Open source S4
mproves performance on both human and robot evaluations. After distillation, our model supports real-time, minute-long rollouts, enabling live VR interaction. Applications in policy evaluation, model-based planning, and world-action modeling further demonstrate the model's utility for embodied decision-making.
Open source S5
Source：arXiv Robotics — research abstracts · arxiv.org