Robot skill learning

摘要: Learning robots that can acquire new motor skills and refine existing ones have been a long standing vision of robotics, artificial intelligence, the cognitive sciences. Early steps towards this goal in 1980s made clear reasoning human insights will not suffice. Instead, hope has offered by rise modern machine learning approaches. However, to date, it becomes increasingly off-the-shelf approaches be adequate for robot skill as these methods often do scale into high-dimensional domains manipulator humanoid nor they fulfill real-time requirement domain. As an alternative, we propose divide generic problem parts well-understood from robotics point view. After designing appropriate basic components, serve ingredients general approach learning. In paper, discuss our recent current progress direction. such, present work on control, elementary movements, well complex tasks. We show several evaluations using both real physically realistic simulations.