Ropedia W0: Scaling Loco-Manipulation World Models across Embodiments with Human Experience
Summary
Ropedia W0 是首个跨人类与多机器人具身的 loco-manipulation 世界基础模型,通过统一运动学表征和 10k 小时人类 + 3k 小时机器人数据实现五个基准上的最先进性能,蒸馏后支持实时分钟级 rollout 与 VR 交互。
View Cached Full Text
Cached at: 10/21/26, 07:01 AM
Paper page - Ropedia W0: Scaling Loco-Manipulation World Models across Embodiments with Human Experience
Source: https://huggingface.co/papers/2610.26855 Authors:
,
,
,
,
,
,
,
,
,
,
,
,
,
Abstract
Ropedia W0 is a world foundation model for cross-embodiment loco-manipulation that unifies human and robot data via kinematic representations and achieves state-of-the-art results across diverse benchmarks.
World models enable embodied agents to anticipate action consequences, yet jointly modeling locomotion and manipulation across diverse embodiments remains challenging due to heterogeneous action spaces and limited multimodal supervision. We present Ropedia W0, the firstegocentric loco-manipulationworld foundation modelspanning humans and robots with diverse embodiments. 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 unifiedkinematic representationencodes locomotion through camera poses and manipulation through projected skeletons, with robot skeletons combining embodiment-specificforward kinematicswith extensive manual keypoint annotation and correction. We also 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 fivecross-embodimentbenchmarks (DreamDojo, Ψ_0, ω-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 improves 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, andworld-action modelingfurther demonstrate the model’s utility for embodied decision-making.
View arXiv pageView PDFProject pageGitHub24Add to collection
Get this paper in your agent:
hf papers read 2610\.26855
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper1
#### ropedia-ai/Ropedia-W0 Image-to-Video• Updated1 day ago • 15
Datasets citing this paper1
#### ropedia-ai/Xp-Teleop Updated1 day ago • 219 • 6
Spaces citing this paper1
Collections including this paper0
No Collection including this paper
Add this paper to acollectionto link it from this page.
Similar Articles
λ_0: Scaling Open-World Egocentric Human Data for Humanoid Loco-Manipulation
该论文提出 HumanVerse——一个500小时的开放世界第一视角人体全身操作数据集,以及基于此训练的人形基础模型 λ_0,实现了无需目标任务机器人演示的零样本迁移到 Unitree G1,在 SIMPLE 上达到 90.0% 成功率。
InterMimicGen: Scaling Humanoid Loco-Manipulation through Self-Evolving Motion Imitation
InterMimicGen 提出一个自我进化的运动模仿框架,通过数据飞轮将稀疏的人-物交互动作捕捉数据转化为持续扩展的、可执行的人形机器人全身 loco-manipulation 动作资源,并实现向真实机器人的迁移。
@seclink: Robot World Models (New Dimension, 0 Deduplication = New Information) Core Projects: - Awesome-WAM (OpenMOSS): Comprehensive Paper List of World Action Models, including DreamDojo (General-Purpose Robot World Model Learned from Human Videos) - awe…
Introduces two projects related to robot world models: Awesome-WAM (OpenMOSS) includes papers such as World Action Models and DreamDojo; awesome-physical-ai curates a collection of papers on VLA models, world models, and embodied foundation models (including NVIDIA Cosmos Predict2.5).
RoboJEPA: Scaling Robotic Latent World Models
RoboJEPA 基于 JEPA 架构,在涵盖 12 种机器人具身的大型真实机器人数据集上训练,首次为多具身机器人世界模型建立了 scaling law:其 latent rollout 的“想象力误差”随计算量呈二阶幂律变化,并可作为真实机器人规划性能的可靠代理。该模型参数量达 8B,是迄今最大的 JEPA 预测模型,支持零样本部署为长时程规划的机器人智能体,并开源全部检查点与训练/部署代码。
GreenWAM: A World Foundation Model for Robotics
GreenWAM 提出一个机器人世界基础模型,通过共享的解剖学动作空间联合训练 20 个机器人家族,使其既能作为世界模型预测动作后果,又能作为策略控制不同形态的机器人,并在多个基准上取得领先表现。