A conceptual framework to support content-based multimedia applications

Abstract

Future applications will require content-based representations of multimedia. This paper elaborates on what is meant by the term "content", when applied to multimedia. The engineering view of content as "the bits in the stream" is contrasted with the linguistic view of content as "meaning". A linguistic model of content is described and expanded to incorporate the needs of digital multimedia. The model is used to clarify existing spatial and temporal segmentation terminology used for multimedia. The importance of the model to standardisation activities, particularly MPEG-7, is discussed. It is argued that the model provides a clear separation of content from the display, transmission and compression technologies used in the creation and distribution of the content. The application of the model is illustrated by discussing its application to an example scenario.