Xiaolei Huang
Lehigh University
Computer Science
and Engineering Department
Date: 05/02/2007
Location: CS Conference Room, NAC 8/207
Time: 12:00 PM ~ 1:00PM
This talk will present a new deformable modeling strategy aimed at integrating shape and appearance in a unified space. Think of traditional deformable models as "active contours" or "evolving curve fronts", the new deformable shape and appearance models we introduce are "deforming disks or volumes". These new models have not only boundary shape but also interior appearance. As a model's shape deforms, its interior appearance statistics are learned online adaptively. A common deformation scheme, the Free Form Deformations (FFD), parameterizes warping deformations of the volumetric space in which the model is embedded in, hence deforming both model boundary and interior simultaneously.
Using the new volumetric deformable models, we propose novel and efficient algorithms that coherently integrate shape and appearance information for robust image segmentation, shape registration, and image registration. We also present several applications of this research on topics including heart modeling and wall motion analysis from tagged MRI images, synthesis and re-targeting of high-resolution 3D facial expressions, and PET/CT multi-modal image registration. Time permitting, I will also talk about a robust point linking algorithm which is a local registration framework that matches visually dissimilar local regions with high accuracy by finding correspondences using only geometric context without comparing the local appearances.
Xiaolei Huang received her doctorate and masters degree in computer science from Rutgers, the State University of New Jersey and her bachelors degree in computer science from Tsinghua University, China. She is currently an Assistant Professor in the Computer Science and Engineering department at Lehigh University, Bethlehem, PA. Her research interests include medical image analysis, computer vision, and machine learning. Her research focuses on developing novel and robust algorithms for segmentation, registration, classification and deformable modeling. In these areas, she has published numerous journal articles, book chapters, and conference proceedings papers. A member of the Institute of Electrical and Electronics Engineers and the Biomedical Engineering Society, Dr. Huang serves as a reviewer for several conferences and journals, including IEEE Transactions on Pattern Analysis and Machine Intelligence, Medical Image Analysis, Machine Vision and Applications, and Graphical Models.
The lecture series is supported by Grove School of Engineering and a NSF grant to establish PRISM (Perceptual Robotics, Intelligent Sensors and Machines) center at CCNY