Research Goals
To understand people, objects, and the world, we try to grasp their appearance, semantics and causality. Our long-term research goal is to enable computers to capture subtle and complex changes through vision, to perceive, understand, and reconstruct the three-dimensional world.
Research Directions
Human-centered visual computing is the primary research direction of our lab. This concept not only emphasizes that humans are the subjects of visual perception but, more importantly, seeks achievable ways for technology to serve humanity and improve human well-being. Human-centered visual computing includes the following specific research areas: human pose estimation, human surface reconstruction, video action understanding, dynamic environment perception, medical image analysis, and visual detection & measurement.
Research Projects
Our lab is dedicated to the research and application of state-of-the-art computer vision technologies. Our researches have been applied in next-generation AI demonstration scenarios such as autonomous driving, intelligent medical treatment, and smart factories. Related research projects include: collaboration on VRU intent recognition and prediction technology, multimodal fine-grained intelligent assessment for early diagnosis of Parkinson's disease, deep learning based research on visual inspection algorithms, and so on.
