首页|Findings from Georgia Institute of Technology Update Understanding of Robotics ( Learning Prehensile Dexterity By Imitating and Emulating State-only Observations )
Findings from Georgia Institute of Technology Update Understanding of Robotics ( Learning Prehensile Dexterity By Imitating and Emulating State-only Observations )
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
2024 OCT 03 (NewsRx)-By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Investigators publish new report on Ro botics. According to news reporting from Atlanta, Georgia, by NewsRx journalists , research stated, "When human acquire physical skills (e.g., tool use) from exp erts, we tend to first learn from merely observing the expert. But this is often insufficient." The news correspondents obtained a quote from the research from the Georgia Inst itute of Technology, "We then engage in practice, where we try to emulate the ex pert and ensure that our actions produce similar effects on our environment. Ins pired by this observation, we introduce Combining IMitation and Emulation for Mo tion Refinement (CIMER) - a two-stage framework to learn dexterous prehensile ma nipulation skills from state-only observations. CIMER's first stage involves imi tation: simultaneously encode the complex interdependent motions of the robot ha nd and the object in a structured dynamical system. This results in a reactive m otion generation policy that provides a reasonable motion prior, but lacks the a bility to reason about contact effects due to the lack of action labels. The sec ond stage involves emulation: learn a motion refinement policy via reinforcement that adjusts the robot hand's motion prior such that the learned object motion is reenacted. CIMER is both task-agnostic (no task-specific reward design or sha ping) and intervention-free (no additional teleoperated or labeled demonstration s)."
AtlantaGeorgiaUnited StatesNorth a nd Central AmericaEmerging TechnologiesMachine LearningRobotRoboticsGe orgia Institute of Technology
NETL
NSTL
