DEERIN is a set of four modular, open-source software prototypes developed through the Intelligent Systems for Screen Archives (ISSA) project. The name stands for Data Enrichment, Exploration, Retrieval, and Interaction — the four functional areas around which the tools are organised. The prototypes are designed for moving image archives that want to experiment with AI technologies at scale and in their own collections, without depending on commercial vendors or requiring specialist in-house expertise from the outset.
Image credit: Miguel Vieira
The four modules
- Data Enrichment — automated annotation and tagging of audiovisual content using computer vision, including semantic segmentation of different types of material
- Exploration — tools for browsing and visualising large collections in ways that surface patterns across time, subject, and form that would be invisible to manual inspection
- Retrieval — geo-search and semantic search interfaces that go beyond keyword matching to enable content-based queries across a collection
- Interaction — interfaces for creative reuse and human-in-the-loop engagement with screen heritage material, including automated audio description for accessibility
Pilots and partners
The prototypes are being piloted by five archives spanning all UK nations through the ISSA project: National Library of Scotland, National Library of Wales, Northern Ireland Screen, North West Film Archive, and Yorkshire Film Archive. These pilots are building cross-institutional knowledge and practical skills around responsible AI use in screen heritage.
All code is open source under an MIT licence: github.com/kingsdigitallab/issa
See Fantastic Futures 2025 for a short video outlining the DEERIN prototypes, anf FrameSense, which is a tool that we have been testing in this stage of early development.