Seminar Speaker
Date Friday, February 11, 2011, 12:30 pm, SB 104 (Auditorium)
Title: Make Street View 3D
Speaker: Ruigang Yang
Abstract
Imagine that you can see anywhere in the world without leaving your room. Google streetview is a popular product that has made this possible. But the view is in 2D and taken from a fixed and sometime unnatural perspective. Having a full 3D model can provide more realistic and immersive experience to the end user. In addition to virtual tourism, the reconstructed 3D model can be used for many other applications such as urban planning and simulation and training.
In my talk, I will first present some results from the Urbanscape project (in collaboration with UNC-Chapel Hill) that generates in real-time detailed 3D models from video images. Then I will show our effort to understand the 3D model better. e.g., clustering similar patterns and attaching semantic information to the raw pixel or point cloud. Finally I will discuss our ongoing project to enhance the quality of 3D data by performing sensor fusion. In the long run, we envision that a fully automatic pipeline that will make creating realistic 3D Urban models as simple as taking pictures.
Biography
Ruigang Yang received the MS degree in Computer Science from Columbia University in 1998 and the PhD degree in Computer Science from the University of North Carolina, Chapel Hill, in 2003. He is an Associate Professor in the Computer Science Department, University of Kentucky. His research interests include computer vision, computer graphics, and multimedia. He is a member of the IEEE Computer Society and the ACM. He is received an NSF Faculty Early Career Development (CAREER) Program Award in 2004.
He has published over 70 papers. These papers, according to Google Scholar, have received over 900 citations. He has been in the program committees in most international conferences in computer vision including ICCV, CVPR, and ECCV. He is an associate editor of the internal Journal of Machine Vision and Applications. This year he is a program co-chair for 3DIMPVT and an area chair for CVPR.