Integrating Audiovisual and Semantic Metadata for Applications in Broadcast Archiving

This paper describes the work done in the PrestoSpace project dealing with audiovisual and semantic content analysis, content description and retrieval in a broadcast archive environment. We present the metadata model that is used in the PrestoSpace system and discuss its design. A key issue is the integration of audiovisual content description and semantic metadata in the model as well as the use of a knowledge base that handles semantic information across the content set. We describe the audiovisual and semantic content analysis tools that are used to extract the metadata represented in the model and discuss the joint use of this information for content segmentation and retrieval.