With a unique combination of the Western canon of video art history since the 1960s and a variety of music videos, the IMAI - Inter Media Art Institute's collection offers important insights into the broad spectrum of the artistic and musical video history. Thanks to ongoing digitization, more than 3,000 works are available for media and art historians. With the increasing availability of time-based material in the archives, a new challenge has arisen: who is going to watch it all?

Artificial intelligence (AI) models enable the systematic structuring of huge amounts of data for the purpose of gaining knowledge from existing information. A process such as machine learning can help to take a new look at existing knowledge in digital archives. The project “IMAI – Inter Media Collection Research with Artificial Intelligence” will investigate these adaptive machines in order to support archivists, art and media historians in examining audiovisual materials and making the handling of audiovisual collections more efficient and convenient.

In the first phase of the project (2024), a test of available prototypes and a review of IMAI's video archive will be used to weigh up application options. The tools used to date for indexing archives will be considered, as will the working methods of other users concerned. In the long term, a software tool will be developed that can automatically analyze and pre-structure video and film archives and be trained and optimized through manual use for the purposes of research, organization, and metadata collection.

The project is funded by the Ministry of Culture and Science of the State of North Rhine-Westphalia (Ministerium für Kultur und Wissenschaft des Landes Nordrhein-Westfalen). The projct is a close collaboration of IMAI with MIREVI (Mixed Reality and Visualization), a working group of the Media Department at Düsseldorf University of Applied Sciences.


Researcher: Dominik Böhnisch