Bringing new life to archived audio-visual content

So many new movies and documentaries relies on historic imagery. Without proper preservation efforts, much of our cinematic history can be lost forever or stuck buried on a dusty archival shelf.

The recent blockbuster, They Shall Not Grow Old, directed by Peter Jackson, was the result of bringing real, 100-year old footage of World War I back to life. The film’s restoration artists turned age-ravaged silent footage into colourised, dimensionalised and speed-corrected imagery with natural movement and appearance that allows viewers into the shoes of those who fought the war. This was no doubt a challenging and time-consuming process for the film makers and researchers.
There is over 30 million film hours (Fh) of analogue audio-visual content stored in European archives.
The AV–Treasure project wants to make the process of finding and using this content much easier. As a result, it will create a valuable resource for the industry, audiences and brands.
Valossa - Project Partner BevariE - Piql AS - 001

Research and Development

Research will focus potential clients’ needs with regards to the most valuable and useful metadata for discovery . In addition , the project will research how crowd-sourcing can inform indexing and enhance footage .  This will inform the development of a workflow to digitise and restore analogue media, with advanced metadata extraction. <br> This process will use AI, machine learning and crowdsourcing. The media will be hosted on a new platform that offers enhanced search capability and access. The platform will also be integrated with piqlConnect, allowing for all content be digitally preserved, ensuring access into the future.


Bringing analogue archives into the digital age

With so much content stored in analogue archives, AV-Treasure will activate the treasure trove into the digital age, by giving direct and easy online access to searchable digital archives so Film Heritage Institutions (FHIs) can monetize their assets, – and at the same have them securely preserved for future use. This applies to content of both local, EU and global significance.

Sophisticated and automated high-speed film digitization and restoration
Application of ML for automated indexation and extraction of rich content and metadata
Integrating and publishing of metadata on an open-source online access platform
Crowdsourcing methodology for further and deeper metadata
New, specific layers of indexation detail, e.g. from “racecar” to also indexing specifics to a level unknown today, e.g. “1964 – Ford GT40 mk.l”
Scientific based CS (as an extension of the AI) for adding further and deeper content and context metadata
Integrating with and publishing of relevant content and metadata on an open-source online access platform, enabling access to, - and supporting monetization of the high-resolution content


Eurostars and the Norwegian and Danish funding bodies are providing financing for this<br>€ 2 Million innovative research and development project.

2 years with completion in March 2022.


Vintage Cloud is a world leader in the technologies required to ensure fast and safe digitisation and restoration of valuable film archives. Its goal is to help our customers quickly and easily monetize their assets by making them more accessible with tools that are cost-effective, innovative and intuitive and designed for the delicate task. Vintage Cloud was founded in 2012, and is headquartered in Copenhagen, Denmark.

Peter Englesson
Managing Director


VIA University College is a leading Danish university of applied science, focusing on areas such as health care, teaching, technology, business, design and media.
In addition to 20,000 full-time students, VIA has more than 20,000 annual participants in further education programs, making VIA the largest institution for further education in Northern Europe.

Flemming Binderup Gammelgaard - VIA University College - AV-treasure

Flemming Binderup Gammelgaard
Senior Consultant


Valossa, founded in 2015 and based in Oulu, Finland, is a deep technology company working on artificial intelligence software for video perception. Valossa AI is a world-leading video recognition framework that enables a completely new generation of video analysis and profiling tools bringing forward new ways of monetizing video.
Their vision is to develop an AI that understands video like a human does.

Mika Rautiainen - Valossa - AV-treasure - cropped for web

Mika Rautiainen