MOSAIC: A Multilingual, multimOdal centraliSed AI-based hub for the European media Content market
November 01, 2024 - April 30, 2026
Multilingual & multimodal AI hub for European media, with a strong focus on interoperable metadata
Project Inquiries
Mike Matton
Koen Renders
Image by Pexels,
About MOSAIC
MOSAIC develops a prototypical European AI-enhanced platform that acts as a central, scalable hub to improve how audiovisual news and media can be discovered and accessed across language borders. The project combines multilingual and multimodal AI capabilities—such as translation and subtitling—with tools for search, categorisation and enrichment to support cross-border reuse and accessibility of media content.
A key ambition is to strengthen a European ecosystem where media assets become more findable, more reusable and more accessible, while enabling future integration through scalable APIs.
A core challenge in cross-border media discovery is interoperability: making sure content and metadata can travel consistently between organisations, systems, and languages. Within MOSAIC, VRT’s contribution focuses on metadata and ontology work, with EBUCorePlus as the central reference point. Specifically, VRT investigates how far EBUCorePlus can function as a practical, future-proof standard to support (1) semantic alignment across partners, (2) richer machine-actionable descriptions for search and discovery, and (3) consistent exchange of metadata across platforms and services.
Funding
MOSAIC is an EU-funded project under Grant Agreement No. LC-03303874.
Consortium
MOSAIC is coordinated by Fincons Group and brings together broadcasters and technology/research partners from multiple EU countries.
Coordinator: Matteo Villa, FINCONS S.p.A
VRT: Mike Maton & Koen Renders
Partners
FINCONS, CCMA (3Cat), Deutsche Welle (DW), RTV SLO, Sky Italy, Pangeanic, Carsa, Infalia, IDC, Universitat Autònoma de Barcelona (UAB) and VRT
