I obtained Ph.D. degree at ECE Department of Georgia Institute of Technology, under the supervision of Prof. Patricio A. Vela. Prior to that, I received my M.S. and B.S. from the Department of EECS, Peking University, in 2013 and 2010 respectively.
I am interested in a variety of challenging problems in computer vision and robot perception. A coarse categorization of my works (click for details):
Visual Odometry/SLAM
LiDAR-Camera fusion
Semantic Mapping / Reconstruction
A couple of places that I am fortunate to work at ...
Research Scientist @Facebook Reality Lab. since Aug. 2019
A snapshot of what we are working on at FRL: LiveMaps, Project Aria.
Computer Vision research intern @MagicLeap. May. 2017 - Aug. 2017
I worked on line-assisted 6DoF pose tracking in low-texture indoor environments.
Computer Vision research intern @Clips team, Google. May. 2016 - Aug. 2016
I worked on quasi-dense visual feature tracking for wide-baseline video stabilization.
PhD @IVALab, Institute for Robotics and Intelligent Machines, ECE Dept., Georgia Tech. Aug. 2013 - Jul. 2019
I work with Prof. Patricio Vela on efficient, accurate Visual SLAM.
Specifically, I am interested in constructing a compact yet accuracy-preserving optimization problem that can be solved efficiently for pose tracking and mapping modules.
The direct applications of my research include: 1) on-board state estimation for small robots such as drones, 2) augmented reality devices.
Research intern @GM Collaborative Research Lab, Robotics Institute, Carnegie Mellon University. Jan. 2013 - May. 2013
I worked with Dr. John M. Dolan on the tasks of segmentation and tracking, using point cloud data from multiple LiDAR sensors mounted on an autonomous vehicle.
Research intern @Intelligent Perception and Urban Computing Lab, the University of Tokyo. Jun. 2012 - Aug. 2012
I worked with Dr. Quanshi Zhang and Dr. Xuan Song on the task of semantic mapping for robot platform working in urban scenarios.
Research assistant @POSS Group, Peking University. Dec. 2008 - Jul. 2013
I worked with Prof. Huijing Zhao and Prof. Hongbin Zha on segmentation, moving object tracking, and semantic mapping using 3D point cloud data.