News

LEGION Project Led by HEAS Member Dominik Hagmann Approved Within the Heritage Science Austria 2.0 Programme

The project LEGION – machine LEarninG-enabled Identification of archaeological Objects in the middle daNube river basin, led by HEAS member Dominik Hagmann as a researcher at the OeAI/OeAW in collaboration with colleagues from the TU Wien Computer Vision Lab, has been recently approved within the Heritage Science Austria 2.0 programme. The project integrates machine learning and archaeological expertise to establish a unified, AI-supported typochronology for Roman common ware from the ancient metropolis of Carnuntum (Austria). Based on a corpus of c. 70,000 digitized profile drawings, LEGION develops a scalable and reproducible framework for the classification, dating, and interpretation of archaeological mass finds, combining eXplainable AI (XAI) and human-in-the-loop (HITL) approaches. By linking morphological and contextual data, the project generates new insights into production, distribution, and settlement dynamics along the UNESCO World Heritage Site “Danube Limes,” while providing an open, FAIR- and CARE-compliant digital research infrastructure for future applications in Heritage Science.

 

Funding

 

ÖAW – Heritage Science Austria 2.0 (Heritage_2024-12_LEGION):

https://www.oeaw.ac.at/foerderungen/foerderprogramme/heritage-science-austria

 

Project webpage:

https://www.oeaw.ac.at/en/oeai/research/historical-archaeology/archaeology-of-the-roman-provinces-in-the-latin-west/legion

Research Areas: