Template for Regular Entry

DEFINITION The widespread search engines, in the professional as well as the personal context, used to work on the basis of textual information associated or extracted from indexed documents. Nowadays, most of the exchanged or stored documents have multimedia content. To reduce the technological gap so that these engines still can work on multimedia content, it is very convenient developing methods capable to generate automatically textual annotations and metadata. These methods will then allow to enrich the upcoming new content or to post-annotate the existing content with additional information extracted automatically if ever this existing content is partly or not annotated.