Visit ShanghaiTech University | 中文 | How to find us
HOME > News and Events > Events
Interferometric Video Analysis for Prelens Tear Film Surface Reconstruction
Date: 2018/5/25             Browse: 30

Speaker:     Dr. Dijia Wu

Time:          14:20—15:10, May 25

Location:    Room 1A-200, SIST Building


In this talk, we will present a novel approach to resolve the sign ambiguity problem of phase demodulation from single interferometry image. The problem is formulated in a binary pairwise energy minimization framework and solved by multigrid iterative quadratic psuedobooleanoptimization algorithm. The proposed method has been successfully applied to reconstruct the surface of the prelens tear film from interferometry videos. 


Dijia Wu received the Ph.D. degree in electrical and computer engineering from Rensselaer Polytechnic Institute in 2010, and the M.Eng. and B.Eng. in electrical engineering from Shanghai Jiaotong University in 1999 and 2002 respectively. He was a staff scientist in Siemens Corporate Research, and senior research engineer in Microsoft and Google. He published papers in premium computer vision and machine learning journals and conferences, such as IEEE TMI, CMIU, CVPR, ICCV, and MICCAI. His work at Siemens regarding artificial intelligence based CT rib analysis and visualization received US 100 R&D award. He is currently R&D VP at United Imaging Intelligence. His research interests include machine learning, computer vision and medical image analysis.

SIST-Seminar 18040