The Virtual Intelligence Center (VIC) focus on developing next-generation computational imaging systems that takes full advantage of both optics and algorithms to benefit various applications such aslight field camera, hyperspectral imaging, biomedical imaging etc.
Upgrade traditional 2D images with high-dimensional light filed as the input and integrating cutting-edge AI technologies like deep learning. Our research aims to implement intelligent vision processing mechanism similar to human eyes, then improve the robustness of computer vision in complicated environment. Our research has broad potential application including automatic drive, intelligent surveillance, medical imaging, etc.
Our research aims at exploring the low-dimensional structure among the huge high dimensional data by leveraging the deep learning and sparse/low-rank models, and use it for solving real computer vision problems, including object recognition, face recognition, action recognition, etc.