Sheet Music Benchmark: Standardized Optical Music Recognition Evaluation
2025
Read Sheet Music Benchmark: Standardized Optical Music Recognition Evaluation (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-10488.txt.
Description
Juan Martinez-Sevilla and co-authors set a standard test for machines that read sheet music in a 2025 paper. Optical music recognition turns printed scores into digital notation, but the field lacked a shared benchmark. Their Sheet Music Benchmark holds six hundred eighty-five pages across many textures, from single lines to piano and quartet scores, all in one standard format.
Beside the data, they propose a scoring metric tuned to music. It extends the usual symbol error rate into a fine-grained count, crediting or blaming separate parts like note heads, beams, pitches, and accidentals. That detail lets users see exactly where a system fails. Standard splits of the data support training and fair comparison. Baseline runs with current methods anchor the scale. The authors offer both pieces as shared ground for the field.
Catalog record
| YEAR | 2025 |
|---|---|
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 28 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-10488