MUSA IST Project

Project Summary

MUSA (2002-2004) aimed at the creation of a multimodal multilingual system that converts audio streams into text transcriptions, generates subtitles from these transcriptions and then translates the subtitles in other languages. MUSA operated in English, French and Greek. A state-of-the-art Speech Recognition system was enhanced and improved to meet the project settings. An innovative Machine Translation scenario combining a Machine Translation engine with a Translation Memory and a Term Substitution module was designed. Sentence condensing for subtitle generation was performed by an automatic analysis of the linguistic structure of the sentence. MUSA combined core speech and language technologies in a real-life application addressing a pan-European audiovisual audience that depends on subtitles to overcome the linguistic barriers.