3D Graphics

The movie industry has already proven that photorealistic computer-generated imagery is able to trick many people into believing the unbelievable. This is not yet attainable when rendering interactive, immersive virtual worlds in real-time, due to this interactivity which (1) limits the rendering time per frame, (2) only allows for rendering one frame at a time, and (3) imposes memory and bandwidth restrictions.

Research into appropriate 3D data representations and compression techniques is needed in order build adaptive virtual worlds where each object is represented at an appropriate quality level using a minimal amount of memory. Such research should not only consider data from a (classical) storage and transmission perspective, but should moreover consider their interactive rendering and, consequently, consider their run-time memory requirements.

Furthermore, the availability of high-quality 3D data, which is becoming more common, changes the priorities of such mesh compression systems. While losslessness remains a valid concern, low-to-mid-range bit rates increase in importance and visually indistinguishable results are obtained at only a fraction of the lossless rate. Higher frequency information becomes apparent only upon very close inspection. Consequently, also ROI-aware mesh coding and random access become increasingly important.

 

  • Paper accepted in IEEE Transactions on Multimedia: Representing and Coding Light Field Images and Video October 14, 2019 Paper accepted in IEEE Transactions on Multimedia: Representing and Coding Light Field Images and Video Our paper, “Steered Mixture-of-Experts for Light Field Images and Video: Representation and Coding“, has been accepted in IEEE Transactions on Multimedia Key observations: Introduction of a novel representation method for any-dimensional image data embedded in a strong Bayesian framework Multiple short conference papers have been presented on the subject, but no full paper was yet published. Extra novelties ...
  • Paper accepted at Computer Graphics Forum: 3D mesh compression July 1, 2016 Paper accepted at Computer Graphics Forum: 3D mesh compressionOur paper, “Scalable Feature-Preserving Irregular Mesh Coding”, has been accepted for publication in Computer Graphics Forum and will be presented at the 2017 Eurographics conference in the session on Geometry Processing. This paper presents a novel wavelet-based transform and coding scheme for irregular meshes. The transform preserves geometric features at lower resolutions by adaptive vertex sampling ...

View all posts