To determine the odometry information from monocular visual information.
To reconstruct a 3D point cloud and camera poses from sequential images.
Concurrent assignment and planning of trajectories for multiple robots (CAPT)
To map the structure of an indoor environment using the information from a walking humanoid, sensors including IMU, Kinect and 2D laser range scanner (LIDAR).
To learn a set of HMM models to recognize different arm motion gestures using IMU sensor readings from gyroscopes and accelerometers.
To estimate the intrinsic parameters and radical lens distortion parameters of a camera using a checkerboard pattern. And recover camera poses with respect to the pattern.
Estimate the camera orientation from raw inertial measurement unit data
Plan a 3D path with the controlling of both a quadrotor in simulation and an actual quadrotor
To detect and estimate the depth of a red barrel using color segmentation and probabilistic models.
Applied machine learning algorithms include SVM, Decision Tree, Naïve Bayes, KNN, Linear regression, PCA
To detect and replace faces in images and videos automatically.
To morph a face from one to another.
Align multiple small images properly and create an image mosaic
Attended the 2013 World Solar Challenge and raced 3000Km across Australia.
To develop a 80Kw Motor controller for HEV and research on permanent magnet synchronous motor control algorithm.
Ranked 2nd in the Chinese University Students’ Creativity Forum
Exhibited in the Beijing Inertial Technology Exhibition
Won The First Prize in the National Students Intelligent Car Competition