论文/arXiv 2610.22340·13 天前

从监督式玩耍中获得灵巧性

DSP用数分钟人类玩耍克隆策略，自主采集数据训练世界模型，并在仿真与真机灵巧操作上验证。

AI 整理

• 本体 SDK 与控制
• 仿真与训练
• 数据与采集

原文链接：Dexterity from Supervised Play（arXiv 2610.22340 · 发布日期：2026-10-15）

Dexterity from Supervised Play

DSP用数分钟人类玩耍克隆策略，自主采集数据训练世界模型，并在仿真与真机灵巧操作上验证。

核心要点

• 面向接触丰富的灵巧操作，核心问题是机器人如何自主收集所需交互经验。
• 先通过行为克隆学习玩耍策略，再自主部署以收集数小时交互数据训练世界模型。
• 推理时玩耍策略为MPC提出候选动作序列，并可在世界模型中自博弈蒸馏目标条件策略。
• 真机上较hindsight行为克隆成功率高69.6%，用少490倍数据达强化学习相当成功率。

验证范围与局限：验证范围为仿真和真实世界中的欠驱动推动及多指手任意到任意手内姿态重定向任务。

英文摘要 (Abstract)

How can robots autonomously collect the contact-rich experience needed to learn dexterous manipulation? We present DSP: Dexterity from supervised play, a framework in which a robot learns a play policy from minutes of human play through behavior cloning, then deploys it autonomously to collect hours of interaction data to train a world model. During inference, the same play policy then guides model-predictive control (MPC) by proposing candidate action sequences for evaluation under the learned world model. To accelerate inference, we also introduce a method for distilling a closed-loop, goal-conditioned policy through self-play within the learned world model. We evaluate this framework on underactuated pushing and any-to-any in-hand pose reorientation with a multifingered hand, in simulation and real world. In real world experiment, our method outperforms hindsight behavior cloning by 69.6% in success rate and achieves comparable success to reinforcement learning with 490 times less interaction data. These results demonstrate that supervised play can autonomously collect interaction data, guide model-based planning, and enable policy learning without any task-specific demonstrations or sim-to-real transfer.

作者：Botao He, Yiming Dou, Yiannis Aloimonos, Jitendra Malik, Akash Sharma, Ruoshi Liu

来源

本页事实对应的公开页面，以原文为准。

1. arXiv 2610.22340：Dexterity from Supervised Play访问 2026-10-21