学术报告第201408期

报告题目:  Graph cut, convex relaxation and continuous max-flow  problems

报 告 人：Xue-Cheng  Tai

报告人单位：Department  of Mathematics, University of Bergen, Norway

邀 请 方：“多谱信息处理技术”国家级重点实验室

报告人简介：

Xue-Cheng Tai教授是挪威卑尔根大学数学系的终身教授，另外他也是挪威奥斯陆大学的兼职教授和挪威CIPR研究中心的兼职人员。Tai教授由于其在科学计算领域的贡献获得第8届“冯康”计算数学奖。研究领域主要包括数值PDE、优化技术、计算机视觉以及图像处理等，他在SIAM J. Sci. Comput.、International Journal of Computer Vision、IEEE Transactions on Image Processing、IEEE Transactions on Visualization and Computer  Graphics、SIAM J. Numer. Anal.等国际顶级杂志以及CVPR、ECCV等国际顶级会议共发表了100多篇文章。担任多个国际会议的大会主席，并多次应邀做大会报告，目前担任Inverse Problems and  Imaging，International Journal of Numerical analysis and  modelling，Numerical Mathematics: Theory, Methods and  Applications，Advances in Numerical  Analysis等多个国际期刊的编辑及执行编辑。

报告摘要: 

In this talk, we will explore the relationship between  graph cut, convex relaxation and some recent continuous max-flow approaches.  Especially, we will see the continuous versus discrete relationship between  them. There are two advantages with these approaches: 1) different fast  numerical algorithms have been used for these approaches; 2) global minimization  can be guaranteed for some nonconvex problems. We shall explore the connections  between some of these algorithms. Image restoration and segmentation will be  used as examples for applications of these algorithms. Extended application of  ROF model, global minimization of the Chan-Vese model and different global  minimization approach for multiphase labeling problems will be presented. This  talk is based on collaborative work with Bae, Boykov, Yuan, Liu and other  collaborators.

时间：2014年5月8号上午9:30—11:00

地点：逸夫科技科12楼报告厅