Department of Computer, Control and Management Engineering

Menlo;">It's ironic that machines today are able to excel at seemingly complex Menlo;">games such as chess and Jeopardy, yet struggle with basic perceptual Menlo;">and motor tasks that we take for granted.  What are the appropriate Menlo;">perceptual, world and motor representations needed to generate robust Menlo;">behaviors in real-time?  A variety of algorithms use low-dimensional Menlo;">dynamical models to simplify information in high-dimensional Menlo;">trajectories.  I will present some recent work on learning Menlo;">low-dimensional reductions, and show examples of how these Menlo;">algorithms can be implemented on humanoid robots.