Perception, Planning, and Motor Control in Machines vs. Animals
It's ironic that machines today are able to excel at seemingly complexgames such as chess and Jeopardy, yet struggle with basic perceptualand motor tasks that we take for granted. What are the appropriateperceptual, world and motor representations needed to generate robustbehaviors in real-time? A variety of algorithms use low-dimensionaldynamical models to simplify information in high-dimensionaltrajectories. I will present some recent work on learninglow-dimensional reductions, and show examples of how thesealgorithms can be implemented on humanoid robots.