by Monica Bordegoni, Giorgio Faconti, Klaus Kansy, Thomas Rist, Panos E. Trahanias and Michael Wilson
Researchers interested in the architecture of multimedia systems and the generation and retrieval of multimedia documents are invited to participate in task 2 of the ERCIM Computer Graphics Network which is receiving support under the CEC Human Capital and Mobility Programme covering travel and meeting expenses.
Task 2 of the ERCIM Computer Graphics Network is looking into different aspects of intelligent multimedia systems. Issues covered by this group include:
Architecture of Multimedia-Multimodal (M4) Systems
The coordination of media and modalities for input and output is a challenge for the architecture of intelligent M4 systems. The impact of mono-modal concepts (e.g., graphics generation) to the broader context of M4 systems covering natural language generation and multimodal interfaces will be considered. Knowledge Representations
Issues in representing knowledge relevant for the analysis of input and the generation of output in multimedia systems will be discussed. The findings of the AI community about knowledge representations (e.g., the EuroKnowledge consortium is looking into the mostly used systems for knowledge representation) will be compared with the requirements of multimedia systems.
Impact of the Computer Graphics Reference Model (CGRM)
The Computer Graphics Reference Model specifies a layer model of the viewing and input pipeline of graphics systems. How does this model relate to typical AI-based systems for the generation of multimedia documents? A systematic survey of approaches to knowledge-based generation systems may answer this question.
Retrieval of Multimedia Documents
Content-based retrieval of multimedia documents is still an open research topic. Queries for MM documents should be expressable by using the media looked for.
Architecture of a multimedia generation system and its relation to the Computer Graphics Reference Model (CGRM)