Research Projects



Learning Humanoid Reaching Tasks in Dynamic Environments


We propose a novel learning-based motion planning algorithm for solving humanoid reaching tasks in dynamic environments.



Iros 2007 Paper

Iros 2007 Presentation (ppt)

Movie (Quicktime MOV 5MB)


Toward Real-Time Object Manipulation in Dynamic Environments


This wrok proposes a method designed to learn grasps from demonstrations and to learn reaching motions by executing tasks in realistic dynamic envionments.



Siggraph 2007 Poster

SIGGRAPH 2007 Presentation