DexRoam: Learning Mobile Bimanual Dexterous Manipulation from Egocentric Whole-Body Human Demonstrations
Overview
Accepted at CoRL 2026, DexRoam is a system developed by researchers from HKUST, the Beijing Institute for Brain Intelligence, and Peking University. It enables robots to learn complex mobile bimanual dexterous manipulation tasks from first-person human demonstrations using a lightweight capture setup with Meta Quest 3 headsets and ZED Mini cameras. The framework employs three-stage alignment (body, semantic, and temporal) to bridge human-robot differences without requiring external motion capture equipment. Experiments show that combining human demonstrations with robot data improves task success rates on GR00T N1.7 and π₀.₅ models across five tasks, allowing robots to learn new manipulation strategies not present in their original training data. The project's code for data collection, alignment, and policy training is open-sourced.
Generated from attributed reports · Updated 2 days ago
Event evidence and corrections
1 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
Data origin: human egocentric
“论文标题：DexRoam: Learning Mobile Bimanual Dexterous Manipulation from Egocentric Whole-Body Human Demonstrations arxiv 链接：https://arxiv.org/abs/2609.35761 项目主页：https://dexroam.github.io/ 采集对齐代码 链接：https://github.com/zhourui9813/DexRoam 策略训练代码链接：https://github.com/zhourui9813/DexRoam-Policy-Training 方法框架：从人类全身示范到机器人策略学习 轻量化采集人类与机器人全身示范 图 1 DexRoam 数据采集系统 DexRoam 构建了一套无需外部动捕设备的便携式第一视角全身采集系统。”
Exact source · revision 1jiqizhixin
Reported quantity · Reported success rate: 29 percent · reported trials · Differing source assertions
“其中，「人类预训练＋人机联合微调」表现最佳，GR00T N1.7 的平均成功率从 29%提升至 56% ，π₀.₅从 32%提升至 57% 。”
Exact source · revision 1jiqizhixin
Reported quantity · Reported success rate: 56 percent · reported trials · Differing source assertions
“其中，「人类预训练＋人机联合微调」表现最佳，GR00T N1.7 的平均成功率从 29%提升至 56% ，π₀.₅从 32%提升至 57% 。”
Exact source · revision 1jiqizhixin
Artifact availability · code: available · Differing source assertions
“论文标题：DexRoam: Learning Mobile Bimanual Dexterous Manipulation from Egocentric Whole-Body Human Demonstrations arxiv 链接：https://arxiv.org/abs/2609.35761 项目主页：https://dexroam.github.io/ 采集对齐代码 链接：https://github.com/zhourui9813/DexRoam 策略训练代码链接：https://github.com/zhourui9813/DexRoam-Policy-Training 方法框架：从人类全身示范到机器人策略学习 轻量化采集人类与机器人全身示范 图 1 DexRoam 数据采集系统 DexRoam 构建了一套无需外部动捕设备的便携式第一视角全身采集系统。”
Exact source · revision 1jiqizhixin
Artifact availability · code: available · Differing source assertions
“论文标题：DexRoam: Learning Mobile Bimanual Dexterous Manipulation from Egocentric Whole-Body Human Demonstrations arxiv 链接：https://arxiv.org/abs/2609.35761 项目主页：https://dexroam.github.io/ 采集对齐代码 链接：https://github.com/zhourui9813/DexRoam 策略训练代码链接：https://github.com/zhourui9813/DexRoam-Policy-Training 方法框架：从人类全身示范到机器人策略学习 轻量化采集人类与机器人全身示范 图 1 DexRoam 数据采集系统 DexRoam 构建了一套无需外部动捕设备的便携式第一视角全身采集系统。”
Exact source · revision 1jiqizhixin
Dataset · embodiments: GR00T N1.7 · Differing source assertions
“团队采用 GR00T N1.7 和π₀.₅ 两个 VLA 模型，并探索三种训练策略： 人机联合训练、人类预训练后使用机器人数据微调，以及人类预训练后进行人机联合微调 ，验证这一框架在不同模型与训练方式下的有效性。”
Exact source · revision 1jiqizhixin
Dataset · embodiments: π₀.₅ · Differing source assertions
“团队采用 GR00T N1.7 和π₀.₅ 两个 VLA 模型，并探索三种训练策略： 人机联合训练、人类预训练后使用机器人数据微调，以及人类预训练后进行人机联合微调 ，验证这一框架在不同模型与训练方式下的有效性。”
Exact source · revision 1jiqizhixin
Report timeline
Follow attributed reports and material updates.
- Machine Heart — Robotics on WeChatSignalDexRoam: Learning Mobile Bimanual Dexterous Manipulation from Egocentric Whole-Body Human Demonstrations
Accepted at CoRL 2026, DexRoam is a system developed by researchers from HKUST, the Beijing Institute for Brain Intelligence, and Peking University. It enables robots to learn complex mobile bimanual dexterous manipulation tasks from first-person human demonstrations using a lightweight capture setup with Meta Quest 3 headsets and ZED Mini cameras. The framework employs three-stage alignment (body, semantic, and temporal) to bridge human-robot differences without requiring external motion capture equipment. Experiments show that combining human demonstrations with robot data improves task success rates on GR00T N1.7 and π₀.₅ models across five tasks, allowing robots to learn new manipulation strategies not present in their original training data. The project's code for data collection, alignment, and policy training is open-sourced.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).