外观
论文arXiv 2606.221744 个月前
OpenHLM开源全身人形VLA模型，少半数数据超越GR00T N1.6
本文由 AI 根据公开内容整理，可能有误，请以原文为准：OpenHLM: An Empirical Recipe for Whole-Body Humanoid Loco-Manipulation（arXiv 2606.22174）
核心内容
针对现有人形机器人系统常将上下肢解耦导致协调能力受限的问题，研究团队发布OpenHLM开源全身人形移动操作方案。该工作通过单变量对照实验验证了构建原生全身VLA模型的技术路径，在跨越较大垂直范围的长程任务测试中，OpenHLM仅使用不到对比基线一半的总演示时间，即实现了对GR00T N1.6和$\Psi_0$两款主流人形VLA模型的性能超越。
技术方案上，研究证实基于关节空间的全身遥操作接口显著优于仅部分暴露自由度的替代方案；在静态及轮式双臂平台预训练的VLA模型能有效迁移至人形全动作空间。此外，引入类UMI的人形数据集HuMI进行异构共训练，使策略无需额外全身遥操即可泛化至新物体与指令。项目已开源代码、训练数据及模型权重，为具身智能提供可复现的全身协同基准。
关键看点
- 全身遥操作：采用关节级全身遥操作接口，相比仅暴露部分自由度的方案，显著提升数据采集质量与下游策略性能
- 跨形态迁移：在静态及轮式双臂平台预训练的VLA模型，可成功迁移至人形机器人全动作空间，降低冷启动成本
- 异构共训练：利用人形版UMI数据集HuMI进行联合训练，无需新增全身遥操数据即可扩展对新物体与新指令的泛化能力
- 数据效率优势：在长程垂直操作任务中，以少于50%的演示数据量，综合性能超越GR00T N1.6与$\Psi_0$等SOTA基线
工程与产业观察
OpenHLM打破了人形机器人上下肢解耦控制的惯性思维，验证了原生全身VLA与跨形态预训练的有效性。其开源的关节级遥操作范式与HuMI共训策略，为解决具身数据稀缺与全身协同难题提供了高价值工程参考，有望加速人形大模型从局部控制向原生全身智能演进。
英文摘要 (Abstract)
Whole-body humanoid loco-manipulation requires coordinating the robot's entire kinematic chain. However, most existing systems typically decouple the upper and lower bodies into separate controllers, limiting such coordination and yielding behaviors similar to those of a wheeled dual-arm platform. In this paper, we ask what it takes to build a whole-body native vision-language-action (VLA) model that maps language and pixels directly to all of the humanoid's degrees of freedom. We conduct a systematic empirical study organized as a roadmap of one-variable-at-a-time experiments across three phases: whole-body teleoperation, VLA model design, and heterogeneous co-training. Our study yields several intriguing findings: a joint-based whole-body teleoperation interface outperforms alternatives that only partially expose the humanoid's degrees of freedom; a VLA pretrained on static and wheeled dual-arm platforms transfers surprisingly well to a humanoid's full action space; and co-training with HuMI, the humanoid analog of UMI, extends the policy to new objects and instructions without additional whole-body teleoperation on those targets. Following this roadmap yields OpenHLM, an open-source recipe for whole-body humanoid loco-manipulation. In a challenging long-horizon task that spans a wide vertical range of the humanoid, OpenHLM outperforms two state-of-the-art humanoid VLA baselines (GR00T N1.6 and $Ψ_0$) using less than half the total demonstration time. Our code, training data, and model checkpoints are available at [https://openhlm-project.github.io/].
作者：Yingdong Hu, Haodong Zhu, Boyuan Zheng, Yihang Hu, Tong Zhang, Zunhao Chen, Junming Zhao, Ruiqian Nai, Yang Gao
来源
本页事实对应的公开页面，以原文为准。