|
Joint NTU, IEEE & PREMIA Seminar Talk by Prof. Yuncai Liu |
|
|
|
|
Wednesday, 04 February 2009 |
 We are pleased to invite you to the following seminar: Human Motion Analysis and Representation in Video Media, by Prof. Yuncai Liu (Distinguished Professor of Changjiang Scholarship Award Program of the Ministry of Education and Honor Professor of the Shanghai Jiao Tong University) on 4 Feb 2009 (Wed). The seminar is jointly organized by School of Electrical & Electronic Engineering, Nanyang Technological University, IEEE Signal Processing Singapore Chapter, and Pattern Recognition and Machine Intelligence Association.
Date: Wednesday, 4 February 2009
Time : 2pm - 3pm
Venue: Meeting Room C (S1-B1c-111), Blk S1, Level B1,
School of Electrical & Electronic Engineering,
Nanyang Technological University.
Speaker:
Professor Yuncai LIU, Distinguished Professor of Changjiang Scholarship Award Program of the Ministry of Education and Honor Professor of the Shanghai Jiao Tong University.
Title: Human Motion Analysis and Representation in Video Media
Abstract:
We present a mode of human motion representation and algorithms of human motion estimation from single and multiple images in video sequences. Inadequately observed information in monocular images and complicated nature of human motion make the 3D human pose reconstruction problem challenging. In order to explore more prior knowledge about human motion, we extract the motion subspace by performing conventional PCA on a small sample set of motion capture data. This allows us to explore the solution space efficiently. We develop an Annealed Genetic Algorithm (AGA) and a Hierarchical Annealed Genetic Algorithm (HAGA) for human motion analysis that search for the optimal solution by utilizing the hierarchical characteristics of the state space. In tracking scenario, we embed the evolutionary mechanism of AGA into the framework of evolution strategy for adapting the local characteristics of the fitness function. We adopt the robust shape contexts descriptor to construct the matching function. Our methods are demonstrated in different types of motions and different image sequences. Results of human motion estimation show that our novel generative method can achieve viewpoint invariant 3D pose reconstruction.
Speaker Bio:
Yuncai Liu is a Distinguished Professor of Changjiang Scholarship. He received a Ph.D. degree from the University of Illinois at Urbana-Champaign, in the Department of Electrical and Computer Science Engineering, in 1990, and worked as an Associate Researcher in the Beckman Institute of Science and Technology, the University of Illinois from 1990 to 1991. Since 1991, he had been a System Consultant and a Chief Consultant of Research in Sumitomo Electric Industries, Ltd., Japan. In 2000, he joined the Shanghai Jiao Tong University as a Distinguished Professor of Changjiang Scholarship Award Program of the Ministry of Education and Honor Professor of the Shanghai Jiao Tong University. He is an associate editor of Pattern Recognition, a council member of China Society of Image and Graphics and a council member of Chinese Transport Engineering Society. Professor Liu is engaged in the wide research fields of Computer Vision and broad areas of Intelligent Transportation Systems (ITS). He made many contributions of original researches in 3D motion estimation and 3D positioning and calibration. Professor Liu jointly published 3 books and 200 papers, conducting the researches of 973 program, 863 program, NSFC and international cooperation. Currently, his research interests are in human motion analysis, robotic surgeries and ITS.
http://www.visionlab.sjtu.edu.cn/Default.htm
|
