外观
论文arXiv 2512.1541110 个月前
MiVLA：人机互模仿预训练VLA模型仿真与真机泛化性能提升
本文由 AI 根据公开内容整理，可能有误，请以原文为准：MiVLA: Towards Generalizable Vision-Language-Action Model with Human-Robot Mutual Imitation Pre-training（arXiv 2512.15411）
核心内容
针对现有视觉-语言-动作（VLA）模型因视角、外观及具身形态差异导致泛化受限的问题，研究团队提出MiVLA模型。该方法通过人机互模仿预训练，挖掘人手与机械臂的行为相似性以构建强行为先验。实验覆盖ARX、PiPer和LocoMan三类异构机器人平台，结果显示MiVLA在仿真环境中性能较π0、π0.5及H-RDT等SOTA模型提升25%，在真实世界机器人控制任务中提升14%，有效融合了人类视频数据的保真度与仿真数据的操作多样性。
技术核心在于利用左右手坐标系下的运动学规则，实现人与机器人动作空间的双向对齐。模型在预训练阶段被要求根据一种具身形式的演示数据，预测该具身的行为轨迹并模仿另一种未见具身的行为，从而建立统一的行为表征。这种双向映射机制解决了相机视角与本体形态不匹配问题，使模型能将海量人类视频中的操作逻辑迁移至机器人端，同时保留仿真环境中的长尾操作技能，显著增强了下游任务的零样本或少样本泛化能力。
关键看点
- 互模仿预训练：基于人手与机械臂行为相似性，通过双向模仿任务将人类数据保真度与仿真数据多样性统一建模
- 动作空间对齐：采用左右手坐标系运动学规则，实现异构具身间动作空间的精确双向映射与语义对齐
- 真机泛化验证：在ARX、PiPer及LocoMan三型平台上实测，真机控制任务成功率较π0等SOTA模型提升14%
- 仿真性能突破：在大规模仿真基准测试中，MiVLA综合表现优于H-RDT及π0.5等主流VLA模型达25%
工程与产业观察
MiVLA验证了基于运动学约束的人机动作空间对齐是解决VLA跨具身泛化的有效路径。对于算法工程师，该方案提供了将海量人类视频转化为机器人可用策略的具体范式；对于产业侧，其在不增加昂贵真机数据采集前提下显著提升异构机器人复用能力，为降低具身智能落地边际成本提供了关键技术支撑。
英文摘要 (Abstract)
While leveraging abundant human videos and simulated robot data poses a scalable solution to the scarcity of real-world robot data, the generalization capability of existing vision-language-action models (VLAs) remains limited by mismatches in camera views, visual appearance, and embodiment morphologies. To overcome this limitation, we propose MiVLA, a generalizable VLA empowered by human-robot mutual imitation pre-training, which leverages inherent behavioral similarity between human hands and robotic arms to build a foundation of strong behavioral priors for both human actions and robotic control. Specifically, our method utilizes kinematic rules with left/right hand coordinate systems for bidirectional alignment between human and robot action spaces. Given human or simulated robot demonstrations, MiVLA is trained to forecast behavior trajectories for one embodiment, and imitate behaviors for another one unseen in the demonstration. Based on this mutual imitation, it integrates the behavioral fidelity of real-world human data with the manipulative diversity of simulated robot data into a unified model, thereby enhancing the generalization capability for downstream tasks. Extensive experiments conducted on both simulation and real-world platforms with three robots (ARX, PiPer and LocoMan), demonstrate that MiVLA achieves strong improved generalization capability, outperforming state-of-the-art VLAs (e.g., $\boldsymbolπ_{0}$, $\boldsymbolπ_{0.5}$ and H-RDT) by 25% in simulation, and 14% in real-world robot control tasks.
作者：Zhenhan Yin, Xuanhan Wang, Jiahao Jiang, Kaiyuan Deng, Pengqi Chen, Shuangle Li, Chong Liu, Xing Xu, Jingkuan Song, Lianli Gao, Heng Tao Shen
来源
本页事实对应的公开页面，以原文为准。