New European project to increase linguistic diversity in audiovisual content
The UAB is involved in the European project MOSAIC (A Multilingual, multimOdal centraliSed AI-based hub for the European media Content market), to create an artificial intelligence-based hub to distribute, access and reuse multimedia content in all European languages, including Catalan.
The UAB is involved in the MOSAIC project thanks to the participation of its TransMedia Catalonia Research Group from the Department of Translation, Interpretation, and East Asian Studies, led by Pilar Orero.
MOSAIC aims to tackle the challenge of language diversity in Europe by developing tools for multilingual translation, automatic subtitling, and AI-driven content adaptation. This includes creating translations for less-resourced languages such as Catalan, Dutch, and Slovenian. With these capabilities, the project seeks to empower broadcasters and media producers to make their content available to a wider audience, enhancing cultural exchange and unity across Europe.
The project will integrate advanced AI-driven functionalities including automated multilingual subtitling, with translation of broadcast content into multiple languages using AI, enhancing accessibility for non-native speakers; AI-enhanced search and navigation to create a user-friendly platform where users can search multimedia content by keywords, visuals, or context; and content adaptation and repurposing, allowing broadcasters to create customised, context-appropriate versions of archival content for new platforms and formats.
The consortium, in addition to the UAB, is made up of ten partners from five EU countries. Led by FINCONS, the broadcasters are CCMA from Catalonia, DW from Germany, RTV SLO from Slovenia, VRT from Belgium and SKY ITALY. Also participating are the SMEs PANGEANIC and CARSA from Spain and INFALIA from Greece.