The AI4Culture project developed a platform which offers access to AI-related resources (including openly labelled datasets for training and testing AI models), alongside a set of deployable and reusable tools and capacity building materials.
The project customised the platform’s components so that they can be reused by cultural heritage institutions in the following scenarios:
- multilingual text recognition in scanned documents
- multilingual subtitles generation and validation
- enrichment with information extracted from images and semantic linking
- machine translation for cultural heritage metadata
Platform components are interoperable with the common European data space for cultural heritage, facilitating data sharing and reusability of cultural content as well as strengthening the connection between the data space and cultural heritage institutions.
Project goals
- Develop an online capacity-building platform offering AI tools, datasets and training materials for the cultural heritage sector.
- Enable cultural heritage institutions to adopt AI technologies to improve metadata, collection management and visibility of their assets.
- Provide openly licensed datasets and deployable AI components to support fine-tuning and reuse of tools tailored to cultural heritage needs.
- Integrate with the common European data space for cultural heritage to promote interoperability, data sharing and reuse across institutions.
Project partners
- Ethnicon Metsovion Polytechnion - Greece
- Europeana Foundation - The Netherlands
- Datoptron - Greece
- Fondazione Bruno Kessler (FBK) - Italy
- Pangeanic - Spain
- Translated - Italy
- Katholieke Universiteit Leuven (KU Leuven) - Belgium
- European Fashion Heritage Association (EFHA) - Italy
- Stichting Nederlands Instituut Voor Beeld En Geluid (NISV) - The Netherlands
- Austrian Institute Of Technology (AIT) - Austria
- Datable - Belgium
- CrossLang - Belgium
Funding information
| Funding stream | The Digital Europe Programme (DIGITAL) (opens in new window) |
| Contract number | LC-01901432 |
| Funders | |
 |