Kaiwen Guo
PhD
郭凯文
Advisor: Qionghai Dai
Broadband Network & Digial Media Lab, Department of Automation,
Tsinghua University, Beijing, China
Email: guokaiwen_neu@126.com


Background
CV
Ph.D.
B.E.
09.2011 ~ 07.2017
09.2007 ~ 07.2011
Department of Automation, Tsinghua University, Beijing, China
Department of Automation, Northeastern University, Shenyang, China
Interests

Real-time Reconstruction of Dynamic Scenes
Real-time Motion Tracking of Body, Face and Hand
GPU Algorithms and Shader Optimization
Photorealistic 3D Modeling
Publications

Real-time Geometry, Albedo and Motion Reconstruction Using a Single RGBD Camera
Kaiwen Guo, Feng Xu, Tao Yu, Xiaoyang Liu, Qionghai Dai and Yebin Liu
ACM Transactions on Graphics 2017 (to be presented at SIGGRAPH 2017)
[Paper]
[Video1]
[Video2]
[Bibtex]
[Project MonoFVV]

Robust Non-rigid Motion Tracking and Surface Reconstruction Using L0 Regularization
Kaiwen Guo, Feng Xu, Yangang Wang, Yebin Liu and Qionghai Dai
IEEE International Conference on Computer Vision 2015
IEEE Transactions on Visualization and Computer Graphics 2017
[Paper'15]
[Paper'17]
[Video1]
[Video2]
[Bibtex]
[Project Page]

Last updated on 09.07.2017
FlyCap: Markerless Motion Capture Using Multiple Autonomous Flying Cameras
Lan Xu, Lu Fang, Wei Cheng, Kaiwen Guo, Guyue Zhou, Qionghai Dai and Yebin Liu
IEEE Transactions on Visualization and Computer Graphics 2017
[Paper]
[Video1]
[Video2]
[Bibtex]
[Project Page]
Awards

2011
Outstanding Graduate with Top 1 GPA, Department of Automation of Northeastern University, Shenyang, China.
2009 ~ 2011
3x National Scholarships, Department of Automation of Northeastern University, Shenyang, China.
2008 ~ 2011
7x First Class Scholarships, Department of Automation of Northeastern University, Shenyang, China.
We propose a real-time method that uses a single-view RGBD input to simultaneously reconstruct a casual scene with a detailed geometry model, surface albedo, per-frame non-rigid motion and per-frame low-frequency lighting, without requiring any template or motion priors.
We present a new motion tracking technique based on a novel L0 regularizer and an efficient iterative solver to robustly reconstruct non-rigid geometries and motions from a single-view consumer-level depth camera.
We present a new generation markerless motion capture technique, the FlyCap system, to capture surface motions of moving characters using multiple flying cameras integrated in autonomous unmanned aerial vehicles (UAV).

BodyFusion: Real-time Capture of Human Motion and Surface Geometry Using a Single Depth Camera
Tao Yu, Kaiwen Guo, Feng Xu, Yuan Dong, Zhaoqi Su, Jianhui Zhao, Jianguo Li, Qionghai Dai and Yebin Liu
IEEE International Conference on Computer Vision 2017
We propose a novel real-time motion tracking and fusion method called BodyFusion that reconstructs non-rigid surface motions of human performers using a single consumer-level depth camera. We take advantage of articulated priors and solve for both of the skeleton and node deformations in a joint optimization.
[Paper]
[Video]
[Project Page]
Teaching

Data Structures and Algorithms (Spring 2016, Tsinghua University)
Teaching Assistant with Dr. Yebin Liu
[Bibtex]