A Proposal for Multimodal Description Specification and Its Application to Media Generation
Kazunari Ito, Masaya Soga, Hiroaki Saito · 2004
This paper proposes a metadata specification that can describe multimodal information collectively. To unite the description about multi-modalities, we expand the attribute of the GDA(Global Document Annotation) tagset which mainly describes syntactic information. This specification enables explicit linkings of various modalities. As an application, we show the process of generating media contents from the proposed metadata which contain dialog and gesture modalities.