Video over InternEt and Wireless (VIEW) Technologies Laboratory
(M. R. Lyu, I. King and T. T. Wong)

The Video over InternEt and Wireless (VIEW) Technologies Laboratory is an R&D centric laboratory founded by three departments (Computer Science and Engineering, Information Engineering, and Systems Engineering and Engineering Management) of the Chinese University of Hong Kong. The Laboratory focuses on new technologies and applications to support delivery of video over Internet with mobile access capabilities. Five major research areas are 1) video information processing and presentation, 2) wireless and mobile applications development, 3) augmented reality for digital entertainment, 4) computational vision and intelligence, and 5) entertainment computing technology.

Video information processing is one of the techniques to cope with video delivery over mobile environment. Information of different channels in the video is extracted and only relevant and important parts are being transmitted to mobile users. Research areas include speech recognition, character recognition, logo recognition, face recognition, video summarization, text summarization, information retrieval, and data visualization techniques.

Developing applications on wireless and mobile environment is a challenging task. The research target is how to author the media data once and can be used on different wireless and mobile platforms. Details research areas include mobile device adaptation model, performance measurements, information representation, and mobile software engineering.

A previous project "A Multilingual Digital Video Content Hub" was supported by Innovation Technology Fund (total funds: $7.8 million) and it aimed at developing techniques for multilingual video content management and delivery over Internet and wireless environment. A prototype called iView was a reference implementation of the proposed open-standard multimedia content hub and exchange frameworks for various applicants. The iView system provides multiple forms of searching capabilities to the users. They can retrieve the information in an easier and faster manner. The system also helps in building the personal video library for families that reduces the tedious video management works significantly. Unlike the traditional video management or indexing system, the i-View system employs the XML technologies to ease the overhead of integrating different searching and indexing functions. Other developers can plug-in their system components easily into the iView system. Besides, the multilingual capabilities of the iView system make it unique to the Hong Kong multilingual cultural environment.

Another project called "Augmented Reality Computing Arena for Digital Entertainment (ARCADE)" is also supported by the Innovation Technology Fund (total funds: $4.48 million). The project had developed a video object tracking engine (VOTE) as a generic software to track 3D marker and human faces in video contents, so that virtual reality image rendering can be created for entertainment and visualization purposes. Industrial applications for this technology include new entertainment experience for personalized video production, content creation for new generation mobile phones, and augmented reality technology for movie and film production.

A project called "Enabling 'Face-to-Face' Entertainment Computing Technology" (EFFECT) is also funded by Innovation Technology Fund (total funds: $0.99 million). We propose to develop new entertainment computing technology and the required software development kit (SDK) for video based face 3D tracking as well as 3D face model and texture capturing to support "Face-to-Face" digital entertainment applications. The aim of the EFFECT SDK is to solve the face manipulation tasks encountered in the "face-to-face" digital entertainment environment. The EFFECT SDK is being designed to cater various needs in digital entertainment development and it includes the functions for i) face 3D tracking in video, ii) 3D face model and texture capturing, iii) 3D face model compression and iv) 3D face model decompress.

The current project is called "Video Processing Toolkit for Visual Perception Technology (VPT2) in Computer Games" and is funded by Innovation Technology Fund (total funds: $0.99 million). We propose to develop a Video Processing Toolkit to facilitate the use of new Visual Perception Technology in computer games and digital entertainment. The two new visual perception technologies being developed are 1) Single User View Dependent Rendering and 2) Tele-Sharing. Single user view dependent rendering visual perception technology will allow the users to perceive a pseudo 3D object being rendered behind the normal display monitor, when the relative viewing direction between the viewer and the monitor is changed. Tele-Sharing visual perception technology will allow the user to share real object with other users to create communal game-playing experience. With the introduction of new visual perception technologies, computer game developers can explore into a new game perception area and introduce new game-playing experience to end-users. The VPT² Toolkit will be a fundamental building block for new entertainment experiences.

The next project to be funded by Innovation Technology Fund (total funds: $0.99 million) is called "Video Capture and View Construction for Volume Computing" (VC³) in Digital Entertainment. We propose to develop a VC³ Toolkit to facilitate the use of Volume Computing Technology in computer game and digital entertainment. VC³ will extend the current View Dependent Rendering (VDR) capability of "Video Processing Toolkit for Visual Perception Technology" (VPT²) from computer generated graphics to synthetic based View Dependent Video (VDV) in VC³. VC³ Toolkit will provide the functions to i) handle the video capture from a video camera array and ii) view construction from the video camera array. With the introduction of the new volume computing technologies, computer game developers can explore into a new game perception area and introduce new game playing experience to end-users. VC³ Toolkit will be a fundamental building block for creating new entertainment experiences.