From Computational Learning and Motor Control Lab

Resources: Grasping and Manipulation


This video shows how knowledge of expected tactile events can be used to correct for erroneous perception. Using a combination of force control and feedback into a movement primitive, the robot can compensate for even very large errors in perception and creates a robust grasping behavior. The associated paper behind this research received the Overall Best Paper Award at IROS 2011.

Retrieved from http://www-clmc.usc.edu/Resources/ResourcesMoviesGraspingAndManipulation
Page last modified on January 12, 2012, at 08:46 PM