Niveau: Supérieur, Doctorat, Bac+8
The Advene Model for Hypervideo Document Engineering Olivier Aubert, Pierre-Antoine Champin, Yannick Prie LIRIS FRE 2672 CNRS - Lyon 1 University 69622 Villeurbanne Cedex 2004-05 Abstract The Advene (Annotate DVd, Exchange on the NEt) project1 is aimed towards communities exchanging discourses (analysis, studies) about au- diovisual documents (e.g. movies) in DVD format. This requires that audiovisual content and hypertext facilities be integrated, thanks to an- notations providing explicit structures on audiovisual streams, upon which hypervideo documents can be engineered. The Advene framework provides models and tools allowing to de- sign and reuse annotations schemas; annotate video streams according to these schemas; generate and create Stream-Time Based (mainly video- centred) or User-Time Based (mainly text-centred) visualisations of the annotations. Schemas (annotation- and relation-types), annotations and relations, queries and views can be clustered and shared in units called packages. Hypervideo documents are generated when needed, both from packages (for annotation and view description) and DVDs (audiovisual streams). This article is mainly dedicated to introducing the various elements of the Advene model, illustrated with a use scenario, and the Advene prototype, composed of an augmented video player with DVD capabilities for Stream-Time Based Visualisation, while user time based visualisation takes place in a standard web browser, thanks to the XML based template language TAL.
- editing annotation
- annotation
- defined video
- list produces
- utbv vocabulary-list
- resulting hypervideo
- annotate video
- schemas