Buzz's main area of research is in the management of multimedia information, in particular, the dynamic capture and storage of the complex media asset collections associated with animation projects. These include models in (mostly) proprietary formats, raster images in a wide variety of formats, vector drawings, videos in various containers and encodings, sound in various formats, code in a variety of scripting languages (some of them proprietary), and text documents.
There is a special focus on dynamically searching and interconnecting existing assets in order to support a high level of reuse - something that is critical given the extremely time-consuming and skilled process of making animated projects. Reuse goes on at all levels, including modeling, animating, applying materials and texturing, and rendering animated projects, as well as in the compositing stage where final video movies are made.
Another focus is on integrating training materials into asset databases, so that artists, animators, programmers, and video/audio editors can train others on how to perform specialized tasks and how to reuse assets they have created. Training materials can consist of desktop videos, audio files, or text. Users of can also request advice and, using Skype, hook up live with experts.
A prototype system is well underway, with two pieces nearing completing. The first is the multimedia, serverized asset database, along with facilities for uploading, tagging, and reusing assets. The second is the XML-based facility for dynamically introducing new forms of media assets and tags into the database; this is particularly critical, given the rapid evolution of animation, audio, video, and graphics technology. The display of information is also XML based, so that the nature of the assets and tags can be closely tied to the visual layout of the interface.