[9-5] Global Structure-from-Motion and Its Application

文章来源:人机交互技术与智能信息处理实验室  |  发布时间:2017-09-01  |  【打印】 【关闭

报告题目:Global Structure-from-Motion and Its Application
活动地点:中科院软件研究所 5号楼4层 中会议室
演讲人: 崔兆鹏 博士 (瑞士 ETH Zurich
Title:Global Structure-from-Motion and Its Application
Speaker:Zhaopeng Cui, Simon Fraser University and ETH Zurich
Time:9:30 am, September 5, 2017
Venue:Middle Meeting Room, Floor 4, Building 5, Institute of Software

       Structure-from-motion (SfM) is a fundamental problem in 3D computer vision, with the aim of recovering camera poses and 3D scene structure simultaneously given a set of 2D images. SfM methods can be broadly divided into incremental and global methods according to their ways to register cameras. Incremental methods register cameras one by one, while global SfM methods solve all cameras simultaneously from all available relative motions. As a result, global SfM has better potential in both reconstruction accuracyand computation efficiency than incremental SfM. In this talk, I will present our latest global SfM system which could deal with all kinds of camera motions and datasets. Then I will present our new robust video alignment algorithm based on global SfM,which can align videos taken at different times with substantially different appearances. At last, I will briefly discuss some other applications with SfM as well as the future work.

       Zhaopeng Cui has just completed his Ph.D. at Simon Fraser University (SFU), and will be a postdoctoral researcher at ETH Zurich in September, 2017.Before joining SFU, he was a Ph.D. candidate at National University of Singapore. He received his B.Sc. (2009) and his M.Sc. (2012) from Xidian University.His research interests are concentrated in 3D computer vision and computer graphics, more specifically, in large-scale image-based 3D reconstruction,video editing, multi-view stereo and SLAM.
      All are welcome!