Linguistic information plays an essential role in the management of multimedia information, as it bears most of the descriptive content associated with more visual information. Depending on the context, it may be seen as the primary content (text illustrated by pictures or videos), as documentary content for multimedia information, or as one among several possible information components in specific contexts such as interactive multimedia applications. Linguistic information can appear in various formats: spoken data in an audio or video sequence, implicit data appearing on an image (caption, tags, etc.) or textual information that may be further presented to the user graphically or via a text to speech processor.
In this context, dealing with multilingual information is crucial to adapting the content to specific user targets. It requires one to consider potential situations where the linguistic information contained in a multimedia sequence is either already conceived in such a way that it can be adapted on the fly to the linguistic needs of the user, or by using an additional process where content should be adapted before presentation.
Finally, there are a wide variety of applications within which multilingual information may appear. MLIF supports development and implementation of a generic framework for dealing with multilingual content: subtitling of video content, dialogue prompts, menus in interactive TV, descriptive information for multimedia scenes, karaoke management, etc. Such information should be considered in the light of the experience of more specialised communities traditionally dealing with multilingual content, namely the translation and localisation industry.
Web site: http://mlif.loria.fr
Authors: Guillaume Rémy, Julien Ducret, Samuel Cruz-Lara
Contact: Samuel Cruz-Lara