Details

REPERTORIUM explores how artificial intelligence can be harnessed to digitise ancient and classical manuscripts, uniting the preservation of Europe’s musical heritage with cutting-edge optical music recognition (OMR) technologies. Thanks to the project, a complete digital archive of the Solesmes collection - over 100,000 sacred chants - has been created, whose images served to train the OMR system. 

The workshop offers a practical introduction to the GABC tool for chant transcription and a musicological exploration of Cantus Index functionalities. Besides the introduction to the approach, we want to discuss how it methodologically/technically works, and what limitations exist. ICCMU, with the collaboration of the Complutense University of Madrid, has supported the University of Alicante in training OMR for medieval notation by transcribing chants into an extended GABC language, and has also assisted the University of Jaén in integrating the resulting transcriptions into the international Cantus Index database. Meanwhile, the University of Alicante develops deep learning models that enable machines not only to read medieval and classical scores, but also to understand how lyrics align with melodies. REPERTORIUM is funded by the European Union.