I am a PhD in the Department of Computer Science and Engineering at the University of Minnesota. My advisor is Prof. Stergios Roumeliotis and I am a member of the Multiple Autonomous Robotic Systems (MARS) Laboratory.
My research focuses on developing 3D navigation and mapping algorithms for mobile devices and autonomous vehicles. In particular, I have been working on efficient large-scale mapping, real-time visual-inertial navigation, sensor intrinsic and extrinsic calibration, and nonlinear-system observability analysis.
Visual-inertial navigation and mapping system on a mobile device. During data collection, the system estimates the user's 3D position and orientation in real time. In the end, a high-quality feature map is created employing a batch-least-squares estimater using all the saved inertial data and feature tracks.
Efficient large-scale cooperative mapping algorithm using multiple visual-inertial datasets from mobile devices. In addition to point features, line features are utilized to improve the orientation estimation accuracy. An interactive visualization of our created feature maps of different buildings and conference sites is available here.
VINS for localization: Localize a wheeled robot within a previously mapped area on a mobile device by processing visual, inertial, and odometry data.
VINS on a smartphone: A real-time visual-inertial navigation system on a Samsung S4 phone (released 2013)