Sub4All aims at the development of a system for automatic monolingual and multilingual subtitling of multimedia content. In the pilot system under development, the project focuses on English and Greek television programmes from the news and politics domain. We investigate the integration of a speech recognition subsystem in a respeaking environment, as well as the development of a module for audio-script alignment. We also develop a subsystem for sentence condensation of transcribed text. Finally, we will integrate a machine translation engine which, combined with translation and terminological memories, will produce translated versions of the automatically generated subtitles in EN and EL.

Sub4All builds a working prototype for subtitle production that will assist human subtitlers in their work. Thus, the project's final deliverable will become a valuable tool for both subtitle companies and multimedia content providers, by reducing subtitle production time and cost. Sub4All supports the European Union's plans for equal access to information for all people, including citizens with disabilities.