Xiaofeng Xie is a Ph.D of pattern recognition at the South China University of Technology (SCUT). His research interests include nonlinear dimensionality reduction, multilinear subpsace learning and graph based semi-supervised learning. He is member of the SCUT center of Brain Computer Interfaces and Brain Information Processing, which develops brain computer interface, robot control and fMRI prcessing.
His current works focus on the motor imagery classification and hyperspectral image classification.
PhD in Pattern Recognition and Intelligence System, 2018
South China University of Technology
MEng in Pattern Recognition and Intelligence System, 2013
South China University of Technology
BSc in Automation, 20011
South China University of Technology
My research project
We proposed some bilinear dimensionality reduction methods and designed related classifiers for motor imagery BCI system. The more details are shown in demo.
The low dimensional embedding is learned from the high dimensional hyperspectral image by a group local Riemannian embedding (GLRE) algorithm.
The fundamental frequency of voice was extraced by the TANDEM-STRAIGHT tools. And the frequency of speech was changed to the singing. The more details are shown in demo.
The proposed Random Walks algorithm was applied to extract the objective from original 2D/3D image. The more details are shown in demo.