Our Work
Projects
AI tools for music score processing, from automated detection to user-guided correction.
AI tools for music score processing, from automated detection to user-guided correction.
AI-Powered Music Score Reader
ToMN uses deep learning to automatically detect and number measures in scanned music scores. The system processes each page in approximately 1.3 seconds and includes a user-guided correction workflow for reviewing results. Currently in controlled institutional beta.
Areas we're researching as natural extensions of the ToMN pipeline. These are explorations, not commitments.
Researching how the detection pipeline could extend into searchable digital archives for institutional score collections.
Combining score structure data with performance recordings for rehearsal insights. Dependent on ToMN reaching production stability.
Investigating how automated score processing could support music education workflows, informed by beta partner feedback.
We're looking for institutional partners and collaborators who work with music scores daily.