Hermes: encouraging media reuse

Audiovisual archives are part of our collective memory. They show how our society has changed, what events mattered to people and how those events were reported. Thanks to the digitisation work of recent years, a vast treasure trove of audio and video has been digitally preserved. But being archived doesn’t automatically mean being accessible: much of this material remains hidden today. What if it were easier to rediscover the past?

Enter Hermes: encouraging media reuse. Together with content partners in the cultural, media, public and creative sectors, we’re making material from the recent and not-so-recent past easier to find, share and reuse.

Challenge

TV series, films, videos, vlogs and podcasts: the rise of digital technology and the affordability of mobile devices have led to unprecedented growth in audiovisual production. But what about material from the not-so-recent past? There are numerous barriers that often keep this content hidden away in archives.

Sometimes rights get in the way. Sometimes there isn’t a good description. And professionals looking to reuse material don’t always have a straightforward way to request or exchange it. In the Hermes project, we’re tackling these barriers together.

Kleurenproeven op papier gebruikt in Zelfportret met groene hoed, by Rik Wouters, collection KMSKA, artinflanders.be, photo: Dominique Provost, public domain.

Our role

This project brings together a wealth of expertise. In collaboration with partners in the cultural, media, public and wider creative sectors, we’re aiming to make (media) easier to reuse – for our content partners, the general public and creative professionals.

Meemoo initiated the project and has a coordinating role. We’ll also be developing solutions, carrying out analysis, experimenting and more. We can build on previous projects exploring areas including AI and metadata and rights.

Approach: three tracks, one goal

In Hermes, we’re working on three fronts at once. Material can only be reused if people are allowed to view it and know what they can do with it, can find it easily and can exchange it without hassle.

Track 1

Making rights clear

Beeld: Turnerin, Heinrich Hamann, collectie Museum of Arts and Crafts, Hamburg, Duitsland, licentie: CC0

Clear rights information

Can you share a clip? Edit it? Publish it? And under what conditions? To give a clear answer, the copyright status of an archival item needs to be recorded accurately.

That’s why we developed a new rights metadata model: a practical guide that helps organisations record rights information consistently and make an informed decision about whether material can also be made available online. Clear wording and visual guidance make the information easy to understand for the people who want to use it. The model is based on international standards and can be shared across sectors.

clearing-rights

Recording instantaneous discs at the Finnish Broadcasting Company in 1940s, 1940, Archives of the Finnish Broadcasting Company Yle, Public Domain. Via Wikimedia Commons.

Making audio and video accessible!

You can already search descriptions of more than 2 million items from our content partners’ collections at thearchives.be. But can the general public explore the audio and video too? Not yet, but we’re keen to change that.

How? Together with our content partners, we’re exploring their varied collections, starting with audio and video that can be made accessible relatively easily. We’ll begin with material that is rights-free or straightforward to clear – for example, where one clearly identified rights holder can give permission. At the same time, we’re looking into what’s possible with collections where the rights situation is more complicated.

Why this is such a challenge

Making audiovisual material widely accessible is no easy task. A single clip often has a whole host of rights holders: a programme maker, broadcaster, actor, presenter or writer. Add music to the mix and the rights puzzle can quickly become very complex – especially when, after many years, it’s no longer clear who owns which rights.

Track 2

Making it easier to exchange archive material

A content partner might want to reuse material from another content partner, but until now there hasn’t been a simple, central way to arrange this. Since June 2026, new features on thearchives.be have made it much easier for our content partners to exchange material. No more separate emails, WeTransfer links or manual file transfers!

But what about a journalist looking for footage for a news report, or a documentary maker who comes across interesting archive material? We’re setting up an initial pilot for these users to explore what exchanging archive material with external users could look like.

Track 3

AI that can keep growing

Recognising faces, identifying voices or automatically creating summaries: AI can help uncover information that might otherwise remain hidden in thousands of hours of audio and video. We’ve already gained plenty of experience through previous projects such as GIVE and Shared AI.

We’re building on that work by:

  • generating new metadata
    We’re doing this with four applications: facial recognition for photographic material, voice recognition, audio classification and automatically generated summaries based on transcripts we previously created ourselves.

  • improving metadata
    AI can add new descriptions on a large scale, but it also makes mistakes. That’s why we’re exploring how users can easily review and edit AI-generated metadata.

  • making new AI techniques quicker to put to work
    Technology is evolving at a rapid pace. And that’s why we’re building our AI infrastructure so that new models and techniques can be added more easily in the future.

Adding descriptions quickly is great, but the results also need to be reliable. And that’s why we don’t simply unleash AI unchecked – we engage with our content partners, weigh sensitivities carefully, and critically review the outcomes. The Flemish Parliament shares more on this.

Don't see a video? Please check your cookie settings so we can show this content to you too.

Edit your cookie preferences here

💡 Much of this AI work happens behind the scenes: the new metadata mainly helps our content partners make their material easier to search. But we’ll soon be sharing something new with the general public! For the first time, thearchives.be will feature AI-generated titles and descriptions for a selection of clips available to the public.

Working together

We’re delivering this project together with our content partners in the media, cultural and public sectors, as well as suppliers and businesses, representatives of rights holders and creative professionals. More than 50 partners are involved in the Hermes project. Who’s involved, and how, varies across the different strands.

Together, we’re taking things step by step towards the same goal – more audio and video that can be seen, found, shared and used again.

Debbie Esmans

Manager Policy, Partnership and Service Offer

Any questions about track 1 or track 2?

Contact Debbie
Matthias Priem

Manager Archiving

Questions about track 3?

Contact Matthias

This project is part of the Flemish Government’s Resilience Recovery Plan.

This page is loading...