Human Demonstrations as Post-Training Data for Any Robot Body: Retargeting the Object, Not the Hand

📄 arXiv: 2610.20351v1 📥 PDF

作者: Nagasrikanth Kallakuri

分类: cs.RO, cs.AI

发布日期: 2026-10-14

备注: 16 pages, 17 figures, 7 tables, 2 algorithms; 56-page supplement as ancillary file. A subset of episodes with their per-robot outcomes: https://playground.dxtr.co


💡 一句话要点

提出人类示范作为机器人后训练数据以解决目标重定向问题

🎯 匹配领域: 支柱九:具身大模型 (Embodied Foundation Models)

关键词: 人类示范 机器人后训练 目标重定向 任务导向训练 仿真评分

📋 核心要点

  1. 现有方法主要依赖于人类示范进行预训练,缺乏将这些数据有效应用于不同机器人身体的能力。
  2. 论文提出通过重定向对象的运动而非手部动作,利用人类示范作为后训练数据,解决机器人任务导向训练问题。
  3. 实验结果显示,在42个捕获样本中,8个机器人完成157对任务,31.2%的成功率,52.2%失败,14.0%拒绝,显示出方法的有效性与局限性。

📝 摘要(中文)

人类示范目前主要用于预训练视觉-语言-动作模型,缺乏将相同数据用于目标体现或任务导向训练的能力,这对机器人在工业和一般环境中的可扩展部署至关重要。通过简单的设备如手机记录的人类示范可以作为后训练数据,适用于与记录者身体结构和能力不同的机器人。该方法的关键在于重定向对象的运动,而非手部动作。捕获的数据提供了无身体的记录:对象的姿态随时间变化、接触间隔以及重建的3D对象。任何机器人在给定URDF和简短描述后,可以根据这些记录进行求解,并在仿真中与示范的对象运动进行评分。

🔬 方法详解

问题定义:论文要解决的问题是如何有效利用人类示范数据进行机器人后训练,尤其是当机器人身体结构与示范者不同的情况下。现有方法在这一点上存在局限,无法灵活适应不同的机器人平台。

核心思路:论文的核心思路是通过重定向对象的运动,而非手部动作,来实现人类示范数据的有效利用。这种方法使得机器人能够在不依赖于示范者身体特征的情况下,学习和执行任务。

技术框架:整体架构包括数据捕获、对象运动重定向、机器人求解和仿真评分四个主要模块。首先,通过简单设备记录人类示范,然后提取对象的运动数据,接着将这些数据应用于不同的机器人进行求解,最后在仿真环境中进行评分。

关键创新:最重要的技术创新点在于将人类示范数据转化为无身体的记录,允许任何机器人在不同的身体结构下进行学习和执行任务。这一方法与现有方法的本质区别在于不再依赖于手部运动的重定向。

关键设计:在参数设置上,使用URDF描述机器人,并通过简短描述进行求解。损失函数和网络结构的具体细节尚未公开,但实验结果表明,拒绝的案例与身体模型的准确性密切相关。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

实验结果显示,在42个捕获样本中,8个机器人完成157对任务,成功率为31.2%,失败率为52.2%,拒绝率为14.0%。这些数据表明,尽管方法具有一定的有效性,但仍存在改进空间,尤其是在不同机器人模型的适应性方面。

🎯 应用场景

该研究的潜在应用领域包括工业机器人、服务机器人以及任何需要灵活适应不同操作环境的自动化系统。通过有效利用人类示范数据,机器人可以更快速地适应新任务,提高工作效率,降低人力成本,具有广泛的实际价值和未来影响。

📄 摘要(原文)

Human demonstrations are today used largely for pretraining vision-language-action models, with little ability to use the same data for target-embodiment or task-oriented training, which scalable deployment of robots in industrial and general environments depends on. A human demonstration recorded on a simple and scalable device such as a mobile phone can serve as post-training data for a robot whose body differs in structure and capability from the human who recorded it. This is made possible not by retargeting the hand's motion but the object's. A capture yields a body-free record: the object's pose over time, its contact intervals, and a reconstructed 3D object with provenance. Any robot, given as a URDF and a short descriptor, is solved against that record and scored in simulation against the demonstrated object motion as ground truth: the human-to-simulation leg of human-to-simulation-to-robot transfer, measured per robot. Each (robot, episode) pair gets one outcome: pass, failure with a named mode, or refusal, with its reason, before any motion. The method runs as a working system over a growing collection; the set analysed here is a sample of it: 42 captures solved on 8 robots, 157 pairs, of which 31.2% pass, 52.2% fail, 14.0% are refused and 2.5% are unauditable (kinematic feasibility, +/-50 mm / +/-15 deg, in simulation). Refusals held up: every pair the checks refuse on stored evidence yet ran anyway failed (13 of 13; 5 on a body whose public model welds its hand). A description file could not be shown to screen a robot: its outcomes were barely more predictable than guessing. No policy has yet been trained on the record. The record decides what a robot may post-train on; whether the policy then works on the real robot is the simulation-to-robot question, answered downstream.