只在关键接触阶段收集数据，实现高精度操作96%成功率
Data and Learning Where it Matters for Contact-Rich Manipulation
- 关键接触段用自动数据+离线强化学习，自由运动段用传统规划
- 仅需2-2.5小时自主数据，平均成功率96%，远超基线55%
- 适合需要高精度的机器人操作任务，尤其对分布外场景鲁棒
端到端训练的模型在高精度任务中常表现脆弱且泛化能力差。我们发现其根源在于数据收集缺乏结构与聚焦。核心思路是仅在接触密集阶段进行密集数据采集，而简单自由空间运动则依赖传统规划。提出自动化数据采集方案结合离线深度强化学习，无需人工遥控或在线策略更新。在四个真实世界任务中，仅使用2至2.5小时自主数据，平均成功率达96%，显著优于最强基线的55%。值得注意的是，在分布外场景中性能依然稳定，而端到端方法在此类情况表现不佳。本工作为接触密集型任务的精准数据采集提供了新路径，推动高成功率精密应用的发展。
原文摘要 · Abstract (English)
Learned policies trained end-to-end on large datasets often remain brittle in high-precision tasks and struggle with generalization. We find that these limitations largely stem from a lack of structure and focus in data collection. Our key insight is to leverage dense data collection only for the critical segment of contact-rich tasks and to rely on traditional planning during simple free-space motion. We propose an automated data-collection scheme in combination with offline deep reinforcement learning for the critical segment of the task, eliminating reliance on a teleoperator's skill and on online policy updates. Across four challenging real-world tasks, using only 2 to 2.5 hours of autonomous data collection, we achieve an average success rate of 96%, compared to the strongest baseline at 55%. Notably, performance remains high in out-of-distribution scenarios where end-to-end approaches struggle. Our results pave the way for targeted data collection for contact-rich tasks and for high success rates in precision applications.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。