Sensory Intelligence: Target Tracking and On-Line Classification

Yuichi Motai, School of Engineering, University of Vermont

Friday, May 23, 2008

Spanos Auditorium

This seminar is part of the Jones Seminars on Science, Technology, and Society series

The speaker will explain on-going projects at University of Vermont, focusing on dynamic framework of sensory intelligence systems in the following two areas: 1) target tracking and 2) on-line classification of target behavior. For target tracking, he has developed Extended Kalman filters for increasing the localization accuracy of the target, using Delta Quaternion and Iterated Multiple Models. For classification of target behavior, he has included incremental aspects in dynamic methodologies; On-line Principal Component Analysis, Incremental Multiple-class Support Vector Machine, and Incremental Decision Tree. The primary goals in these learning algorithms are to design intelligent systems that adapt to changes in target environments in a dynamic manner. These newly proposed algorithms for video streams are applied to many other domains; such as medical image database (joint work at the Harvard Medical School, Boston, MA) and satellite video streaming (with Air Force Research Lab at Hanscom, MA). He will specifically present how his studies fit well with medical applications more details at his new University.

About the Speaker

Dr. Yuichi Motai has been an Assistant Professor in the School of Engineering at the University of Vermont. He recently accepted a faculty position in Electrical and Computer Engineering at Virginia Commonwealth University, and in the process will launch a new research lab, called Sensory Intelligence Lab. His research interests are in the broad area of computational intelligence and applications for medicine; including medical imaging, computer vision, human-computer interaction, and sensor-based robotics. He earned his B.Eng. in Instrumentation Engineering at Keio University in 1991, his M.Eng. in Applied Systems Sciences at Kyoto University in 1993, and his Ph.D. in Electrical and Computer Engineering at Purdue University in 2002. He was a visiting assistant professor at Massachusetts General Hospital and Harvard Medical School in 2006. He has published more than 30 papers in peer-reviewed journals and competitive conference proceedings in the last 4 years.