RAPID: Robot Agentic Programming from Demonstrations
AI Digest
机器人代理编程演示学习关系程序表示非prehensile操作
RAPID通过单次视觉演示自动生成机器人程序，结合任务规范、动作原始和交互环境，实现程序验证与优化，展现跨场景泛化能力。
RAPID generates robot programs from single demonstrations using task specs, action primitives, and interactive environments, showing strong generalization across object and environment variations.
Key points
- 基于视觉演示自动生成验证机器人程序 Generates verified robot programs from single visual demonstrations
- 采用对象中心关系表示替代运动轨迹 Uses object-centric relational representation instead of motion trajectories
- 在模拟与真实机械臂完成8类非prehensile任务 Completes 8 non-prehensile tasks in simulation and real robots
- 实现物体姿态/材质/环境的跨场景泛化 Generalizes across object poses, materials, and environments
Takeaway: 视觉演示驱动的程序生成为机器人自主操作提供新范式。 / Visual demonstration-driven programming offers new paradigm for autonomous robotic manipulation.
Why it matters 方法将演示学习与程序生成结合，为通用机器人任务提供可复用解决方案。
View original ↗ Back to hot list
This page is an aggregated digest from arxiv; content and hot-score data come from public sources. Copyright belongs to the original authors. We link to originals with nofollow and never republish full text.