MessyMem：移动操作中的做中学记忆
MessyMem: Learning-from-Doing Memory for Mobile Manipulation
浏览论文内容
中文总结 AI 辅助
MessyMem通过持久记忆系统，利用空间3D场景图与交互知识，使移动操作机器人跨任务复用经验，在25任务仿真中任务进度达80%，显著优于基线。
中文摘要 AI 辅助
部署在多个房间并多次访问的移动操作机器人应能通过经验改进：在发现橱柜被锁或从抽屉中找到物体后，机器人应复用该知识，而不是从零开始每个任务。然而，当今的机器人常将每个任务视为新任务：紧凑的场景表示省略了交互派生的知识，原始视频历史难以查询，VLM规划器在推理时进行推理而不持续更新机器人所知。我们提出MessyMem，一种持久记忆系统，使移动操作机器人能从经验中学习并在未来任务中复用该知识。它维护一个空间锚定的3D场景图，包含物体和位置，通过交互学习到的属性和结果对其进行增强，并链接视觉观察以实现细粒度回忆。我们在仿真和真实移动操作机器人上评估MessyMem。在跨越3小时以上的连续25任务仿真中，MessyMem达到80.0%的任务进度，超过最强消融14.8个百分点，超过最强外部基线28.9个百分点，同时从数千个存储的关键帧和超过一小时的历史中检索任务相关证据。
英文摘要
Mobile manipulators deployed across many rooms and visits should improve with experience: after discovering that a cabinet is locked or finding an object in a drawer, the robot should reuse that knowledge rather than start each task from scratch. Yet today's robots often treat each task as new: compact scene representations omit interaction-derived knowledge, raw video histories are difficult to query, and VLM planners reason at inference time without persistently updating what the robot knows. We present MessyMem, a persistent memory system that enables mobile manipulators to learn from experience and reuse that knowledge across future tasks. It maintains a spatially grounded 3D scene graph of objects and locations, augments it with properties and outcomes learned through interaction, and links visual observations for fine-grained recall. We evaluate MessyMem in simulation and on a real mobile manipulator. In a continuous 25-task simulation spanning over 3 hours, MessyMem achieves 80.0% task progress, outperforming the strongest ablation by 14.8 percentage points and the strongest external baseline by 28.9 points, while retrieving task-relevant evidence from thousands of stored keyframes and over an hour into the past.
发表机构
- Stanford University（斯坦福大学）
机构由 AI 辅助整理，请以论文原文为准。
