# OneDemo：从单个演示进行统一策略学习与数据生成

> 原文链接：https://robotics.hellobike.com/papers/2026-10-13-2610-18188
> 频道：papers | 更新日期：2026-10-13 | 标签：arxiv, cs.RO
> 核心摘要：OneDemo将单个成功演示编辑建模为强化学习，用于策略学习和数据生成，并在Meta-World与DexMimicGen任务上验证。

# OneDemo：从单个演示进行统一策略学习与数据生成

::: v-pre
> 原文链接：[OneDemo: Unified Policy Learning and Data Generation from a Single Demonstration](https://arxiv.org/abs/2610.18188)（arXiv 2610.18188 · 发布日期：2026-10-13）

**OneDemo: Unified Policy Learning and Data Generation from a Single Demonstration**

OneDemo将单个成功演示编辑建模为强化学习，用于策略学习和数据生成，并在Meta-World与DexMimicGen任务上验证。

## 核心要点

- 收集多样高质量专家演示成本高，单演示扩增常需手工规则或复杂规划。
- OneDemo将演示编辑表述为结构化强化学习，从稀疏成功奖励学习切换与几何适配。
- 在15个Meta-World任务中，平均53.7k次环境交互内达到92.7%成功率。
- 在9个DexMimicGen任务中生成演示成功率98.4%，训练的BC-RNN策略在同等数据规模下达99.6%。

**验证范围与局限**：验证范围为15个Meta-World任务和9个DexMimicGen任务。

## 英文摘要 (Abstract)

Learning robot policies from demonstrations is often limited by the cost of collecting diverse, high-quality expert data. While an appealing alternative is to expand a single successful demonstration into a diverse set of successful trajectories, existing augmentation methods typically rely on manually designed transformation rules or complex trajectory planning, requiring substantial task-specific engineering and human effort. We introduce OneDemo, a reinforcement learning framework that learns to edit a single successful demonstration for new task configurations. We formulate demonstration editing as a structured reinforcement learning problem, in which an editor learns, from sparse task-success rewards, both when to transition between object-centric trajectory segments and how to geometrically adapt each segment. The learned editor serves both as a task policy and as a reusable demonstration generator. Across 15 Meta-World tasks, OneDemo reaches a 92.7% success rate within 53.7k environment interactions on average. Across 9 DexMimicGen tasks, it generates successful demonstrations with a 98.4% average success rate, and BC-RNN policies trained on the generated data achieve 99.6% average success at matched dataset scale. These results show that learning to edit a single successful demonstration provides a structured and effective basis for both policy learning and scalable demonstration generation.

作者：Ying Qian, Jianshu Hu, Han Fang, Shuguang Li, Yutong Ban
:::

## 原始参考来源 (Sources)

- [arXiv 2610.18188：OneDemo: Unified Policy Learning and Data Generation from a Single Demonstration](https://arxiv.org/abs/2610.18188)