首页>
外国专利>
Device for automatic annotation of e.g. multi-media document, for online recognition of e.g. actors, in video film, has classification test module applying classification text to document to provide annotation of document at output
Device for automatic annotation of e.g. multi-media document, for online recognition of e.g. actors, in video film, has classification test module applying classification text to document to provide annotation of document at output
The device has an acquisition unit for acquisition of an ontology (180) in vicinity of initial instant, and an ontology reduction module (150) reducing the ontology by analysis of text and metadata of multi-media document (160). An ontology settlement module (140) searches textual terms in a database of not-textual documents. A learning module (110) allows learning to take classification module (111) of instances of objects in document. A classification test module (130) applies classification text to multimedia document by using model to provide an annotation (131) of the document at output.
展开▼