We study the embedding of cryptography, such as homeomorphic encryption and garbled circuits, into machine learning algorithms in order to reduce the risk of exposing confidential personal information.
In collaboration with Princeton University and Iowa State University, we study how to continuously and unintrusively authenticate user's identity based on his/her behavior biometrics.
We study the cost-effectiveness and robustness issues in kernel-based machine learning.
Scene text recognition, image dehazing, pose estimation, panorama saliency detection, medical image, air pollution prediction.
Optimize fan performance using Reinforcement Learning (RL).