发表机构
MARS Lab, Nanyang Technological University; Georgia Institute of Technology（南洋理工大学MARS实验室; 佐治亚理工学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究针对机器人功能工具使用泛化问题，提出FORGE两阶段策略，通过关键点轨迹推理解耦功能推理与动作执行，在七工具击打基准测试中，对未见工具表现优于现有方法，平均成功率显著提升。
AI 中文摘要
人类能轻松地将书本、石头或鞋子用于敲钉子，但在特定工具上训练的机器人却无法将相同功能转移到新工具上，这一差距被定义为功能泛化。此类工具虽有视觉上可识别的共同功能意图，但感知相似性并未延伸至动作空间，每个工具都需要完全不同的运动模式。为弥合这一差距，我们探索了包括可供性图像、人类视频提示和二维关键点轨迹在内的中间表示形式，发现关键点轨迹在功能表现力和动作可落地性之间达到了最佳平衡。在此基础上，我们提出了功能推理与基础执行（FORGE），这是一种两阶段策略，将功能推理与动作执行解耦：从无动作数据中预测可泛化的关键点轨迹，然后通过有限的演示将其转化为机器人动作。在七工具击打功能基准测试中，FORGE在模拟和现实世界中对未见工具的表现始终优于现有方法，平均成功率提高了两倍多。
英文摘要
While humans readily repurpose a book, a stone, or a shoe to drive a nail, robots trained on specific tools fail to transfer the same function to novel ones -- a gap we formalize as functional generalization. Functionally equivalent tools share visually recognizable functional intent, such as where contact can occur and how a contact region should move to the target. However, this perceptual similarity does not directly carry over to action space, where each tool demands a different motor pattern to realize the function. To bridge this gap, we explore intermediate representations including affordance images, human video prompts, functional videos and object masks, and 2D keypoint trajectories, finding that keypoint trajectories best balance functional expressiveness and action groundability. Building on this, we present FuncBridge, a two-stage framework that decouples functional reasoning from action execution: learning to predict generalizable keypoint trajectories from action-free data, then grounding them into robot actions with limited demonstrations. Across a benchmark spanning ten tools and three functions, including hitting, sweeping, and hooking, FuncBridge consistently outperforms state-of-the-art methods on unseen tools in both simulation and the real world.
Comments19 pages, 12 figures, 6 tables
