RTVE

RTVE’s Documentary Archive manages millions of audiovisual, audio, and textual assets that are part of Spain’s historical heritage. However, a portion of this archive lacked sufficient metadata, making it difficult to retrieve and reuse in current productions. At the same time, demand for archive content was growing with the proliferation of new programs and video-on-demand platforms. RTVE needed to modernize its cataloging processes with a system that ensured continuity, quality, and scalability.

RTVE implemented an automatic metadata service based on Artificial Intelligence, fully integrated into its ARCA document management system. The solution was deployed in two phases: an initial configuration and testing phase, aimed at ensuring quality and the integration with existing workflows, and a second phase of continuous service dedicated to the massive processing of content.

The system incorporates advanced audio and video analysis technologies: automatic transcription, entity and personality recognition, object, logo, and scene detection, on-screen character extraction, and automatic translation. All of this is combined with a flexible architecture that allows the selection of the best tools available at any given time, ensuring that RTVE remains at the technological forefront.
In addition, an interface was developed to enable documentalists to easily validate the results, combining automation with the professional judgment of the teams. This approach, together with training provided in the use of AI, facilitated the adoption of the solution and eased the implementation of the service within archival processes.

Thanks to this initiative, RTVE has achieved:

• Increased accessibility and reuse of processed material from its documentary archive.
• Facilitated the cataloging of heritage materials, both from the film archive and daily production.
• Optimized time and resources through automated processes that ensure continuity and quality.
• Enabled the creation of new productions requiring archive material, enhancing the use of previously untapped resources.

The implemented solution has strengthened the automatic metadata model that RTVE has been developing in recent years. The service has been properly integrated into our workflows, respecting established timelines and quality metrics, and facilitating the validation tasks of the documental teams. This project has contributed to improving the accessibility of the Archive and reinforcing its operational continuity, maintaining a balance between automation and expert supervision, which is considered essential for the final quality of the metadata.