arXiv 2512.246532026-02-27cs.RO

RoboMIND 2.0：一种多模态、双臂移动操作数据集，用于通用具身智能

RoboMIND 2.0: A Multimodal, Bimanual Mobile Manipulation Dataset for Generalizable Embodied Intelligence

Beijing Innovation Center of Humanoid Robotics（北京人形机器人创新中心）；Peking University（北京大学）

Chengkai Hou, Kun Wu, Jiaming Liu, Zhengping Che, Di Wu, Fei Liao, Guangrun Li, Jingyang He, Qiuxuan Feng, Zhao Jin, Chenyang Gu, Zhuoyang Liu, Nuowei Han, Xian… 展开作者

Chengkai Hou, Kun Wu, Jiaming Liu, Zhengping Che, Di Wu, Fei Liao, Guangrun Li, Jingyang He, Qiuxuan Feng, Zhao Jin, Chenyang Gu, Zhuoyang Liu, Nuowei Han, Xiangju Mi, Yaoxu Lv, Yankai Fu, Gaole Dai, Langzhe Gu, Tao Li, Yuheng Zhang, Yixue Zhang, Xinhua Wang, Shichao Fan, Meng Li, Zhen Zhao, Ning Liu, Zhiyuan Xu, Pei Ren, Junjie Ji, Haonan Liu, Kuan Cheng, Shanghang Zhang, Jian Tang

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

RoboMIND 2.0通过多模态双臂移动操作数据集和MIND-2系统，提升通用具身智能的泛化能力。

摘要：

尽管数据驱动的模仿学习已革新了机器人操作，但当前方法仍然受到大规模、多样化的现实世界演示稀缺的限制。因此，现有模型在长时间双臂任务和无结构环境中的移动操作泛化能力仍然有限。为了弥合这一差距，我们提出了RoboMIND 2.0，一个包含超过31万条双臂操作轨迹的全面现实世界数据集，涵盖六个不同的机器人形态和739个复杂任务。关键的是，为了支持接触丰富和空间扩展任务的研究，该数据集包含了1.2万条触觉增强的片段和2.0万条移动操作轨迹。此外，我们构建了高保真的数字双胞胎，释放了额外的2.0万轨迹模拟数据集，以促进鲁棒的仿真到现实转移。为了充分利用RoboMIND 2.0的潜力，我们提出了MIND-2系统，一个通过离线强化学习优化的分层双系统框架。MIND-2集成了一个高级语义规划器（MIND-2-VLM）以将抽象的自然语言指令分解为地面化的子目标，并结合一个低级的视觉-语言-动作执行器（MIND-2-VLA），生成精确的、具有本体感觉的运动动作。

英文摘要：

While data-driven imitation learning has revolutionized robotic manipulation, current approaches remain constrained by the scarcity of large-scale, diverse real-world demonstrations. Consequently, the ability of existing models to generalize across long-horizon bimanual tasks and mobile manipulation in unstructured environments remains limited. To bridge this gap, we present RoboMIND 2.0, a comprehensive real-world dataset comprising over 310K dual-arm manipulation trajectories collected across six distinct robot embodiments and 739 complex tasks. Crucially, to support research in contact-rich and spatially extended tasks, the dataset incorporates 12K tactile-enhanced episodes and 20K mobile manipulation trajectories. Complementing this physical data, we construct high-fidelity digital twins of our real-world environments, releasing an additional 20K-trajectory simulated dataset to facilitate robust sim-to-real transfer. To fully exploit the potential of RoboMIND 2.0, we propose MIND-2 system, a hierarchical dual-system frame-work optimized via offline reinforcement learning. MIND-2 integrates a high-level semantic planner (MIND-2-VLM) to decompose abstract natural language instructions into grounded subgoals, coupled with a low-level Vision-Language-Action executor (MIND-2-VLA), which generates precise, proprioception-aware motor actions.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。