Computer Graphics

Display of large or multiple movies in arbitrary 3D environments

Both T_Visionarium and Conversations demand the playback of large amounts of video data in 3D worlds. With the aim of solving two problems with one innovation, a system for displaying arbitrarily sized video data on any textured surface in a 3D world will be developed. Key to solution will be:

  • Automatic detection/prediction of the visible regions of the video streams in the forthcoming frame
  • Optimal handling of panoramic and spherical video sources
  • Automatic resolution control to maintain the fixed frame-rate
  • Adoption of a suitable compression schemes
  • Optimal use of the latest generation programmable graphics cards
  • Development of a stand-alone tool for preparation of the video data

Prerequisites: Computer graphics. Higher level shading languages desirable.

Novel methods for combining film and 3D graphics

Building on the ability to display video on any texture and video analysis, novel techniques for mixing 3D graphics and video will be developed. Research will involve addressing the problem of constructing convincing 2D / 3D sets as well as exploitation of video content analysis techniques (ie foreground, background, motion, colour, human forms). This work will build on concurrent work at iCinema in automated video analysis and computer graphics and is expected to involve development for the latest generation programmable graphics cards.

Prerequisites: Computer graphics, image analysis. Higher level shading languages desirable.

Novel rendering methods

The latest generation of programmable graphics cards have made it possible to veer away from the tired ‘texture and lighting’ style adopted for most 3D environments. This project involves research and development into unique and aesthetically pleasing real-time rendering techniques for interactive cinema. The latests graphics libraries, OpenGL 2.0 or DirectX9 are expected to be used.

Prerequisites: Computer graphics. Higher level shading languages desirable.

Facial animation / Lip-syncing

Facial animation is both an important and difficult issue in the production of virtual characters. Its importance and difficulty stem from the same reason: human beings are exceptional experts at the perception and interpretation of facial expressions.
Research opportunites exist in the fields of non-verbal facial animation and facial animation for use with a speech synthesiser. Emphasis will be placed not on photorealistic rendering, but on determining which visual cues are important for meaningful and expressive communication.

Prerequisites: Computer graphics.

Advanced real-time lighting: shadows, reflections, radiosity

Advanced graphics techniques for VR on PC clusters

back to Postgraduate Research Areas