In the past, we focused much of our efforts to develop techniques to support real-time graphics and virtual reality applications. We aimed at developing a distributed virtual environment (DVE) that could support a large number of users interacting with each other. One application of DVEs that we were particularly interested at was online gaming.
In recent years, we have shifted our research interests to Computer Graphics and
Computer Vision. We are particularly interested at developing AI techniques to
support graphics and vision tasks. In computer graphics research, our main
research focus is to develop tools to synthesize various types of graphics
designs. In computer vision, our main research focuses are on detecting and
Our main projects on computer graphics and computer vision include:
(with papers in ACM SIGGRAPH, ACM SIGGRAPH Asia, IEEE
CVPR, ACM SIGCHI)
Project Aim: This research is to develop AI tools for automatic layout
synthesis of different types of graphic designs.
(with papers in ACM SIGGRAPH, ACM SIGGRAPH Asia, ACM Multimedia)
Project Aim: This research is to answer one key question: Can computer design?
(with papers in IEEE CVPR, IEEE ECCV, IJCV, IEEE
Project Aim: This research is to find out the factors that capture the
attention of the human visual system.
and Glass Detection
(with papers in IEEE CVPR, IEEE ICCV)
Project Aim: This research is to study the properties of mirror and glass,
and to develop AI models to automatically detect them.
(with papers in IEEE CVPR)
Project Aim: This research is to study the properties of under-exposed (as
well as over-exposed) images, and to develop AI tools to handle various
types of under-/over-exposure problems.
and Crowd Tracking/Animation
(with papers in IEEE CVPR, ACM SIGGAPH Asia, IEEE TIP, IEEE TCSVT)
Project Aim: This research is to detect and track object (in particular
human) motions for applications such as animation.
(with papers in ACM Multimedia, ACM VRST,
Project Aim: This research is to develop tools to support 3D image
Geometry Model Analysis and Retrieval
(with papers in IEEE TPAMI, IEEE TVCG)
Project Aim: This research analyses 3D geometry models and attempts to
extract useful features that can be used to differentiate deformable
geometry models, for applications such as retrieval.
Last updated in June 2020