UNIVERSITY OF HERTFORDSHIRE COMPUTER SCIENCE RESEARCH COLLOQUIUM presents "Towards Incremental, On-line Learning of Full Body Motion Primitives for Humanoid Robots" Dr. Dana Kulic (University of Tokyo, Japan) 5 August 2008 (Tuesday) Seminar Room D120 Hatfield, College Lane Campus 3 - 4 pm Coffee/tea and biscuits will be available. Everyone is Welcome to Attend Abstract: In this talk, I will first briefly overview the research activities of the Nakamura-Yamane Lab at the University of Tokyo. The talk will focus on our recent work on learning by observation for humanoid robots. Learning behavior and motion primitives from observation is a key skill for humanoid robots, enabling humanoids to take advantage of their similar body structure to humans. The goal of our research is to develop robots which can learn during continuous co-location and interaction with humans in the human environment. In the proposed approach, human motion patterns are abstracted into a dynamic stochastic model, which can be used for both subsequent motion recognition and motion generation, analogous to the mirror neuron hypothesis in primates. An algorithm for on-line, incremental learning of whole body motion primitives by observation of human motion will be described. As new motion patterns are observed, they are stochastically segmented and incrementally grouped together using local clustering based on relative distances in the model space. Over time, the algorithm incrementally forms a tree structure representing the motion space, with specialized motions at the tree leaves, and generalized motions closer to the root. The talk will conclude with an overview of preliminary experimental results and a discussion of future research directions. About the speaker: Dana Kulic received a combined B. A. Sc. and M. Eng. degree in electro-mechanical engineering, and Ph. D. degree in mechanical engineering from the University of British Columbia, Canada, in 1998 and 2005, respectively. She is currently a Japan Society for the the Promotion of Science post-doctoral fellow at the Nakamura-Yamane Laboratory in the Department of Mechano-Informatics at the University of Tokyo, Japan. Her research interests include human-robot interaction, robot learning and humanoid robotics. -------------------------------------------------- Hertfordshire Computer Science Research Colloquium http://homepages.feis.herts.ac.uk/~nehaniv/colloq