Metraipolis › archive › dc_8ccda945da63

Randomized Scores and Diverse Timbres: Augmenting Automatic Music Transcription with Online-Generated Data — arXiv:2610.11197

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.11197
License (as archived)
CC0-1.0
Archived
2026-10-09T07:14:06.761Z
SHA-256 of the archived text
efd0675992c3a1ed0ce712bb28155c8a160f946446d35ad40a743f5ab1a5f77b
Stable URL of this page
https://metraipolis.com/documento/dc_8ccda945da63

How to cite

Randomized Scores and Diverse Timbres: Augmenting Automatic Music Transcription with Online-Generated Data — arXiv:2610.11197 — arXiv (en), https://arxiv.org/abs/2610.11197 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_8ccda945da63 (Metraipolis archive dc_8ccda945da63, sha256 efd0675992c3a1ed0ce712bb28155c8a160f946446d35ad40a743f5ab1a5f77b)
BibTeX
@misc{dc8ccda945da63,
  title = {Randomized Scores and Diverse Timbres: Augmenting Automatic Music Transcription with Online-Generated Data — arXiv:2610.11197},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.11197},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_8ccda945da63; sha256 efd0675992c3a1ed0ce712bb28155c8a160f946446d35ad40a743f5ab1a5f77b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.11197. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Automatic music transcription (AMT) is limited by the scarcity of audio recordings paired with precise symbolic annotations. Synthetic data can provide supervision at scale, but it remains unclear whether effective transfer depends on realistic score structure or broad timbral coverage. We study these factors separately through an online sampler--renderer pipeline. A unified corruption sampler ranges from unmodified MIDI clips through partial corruption to deeply randomized note-event distributions. The renderer converts these events to audio while independently controlling instrument and timbral coverage. A fixed transcription model is trained jointly on offline recordings and newly rendered examples. Controlled ablations reveal an asymmetry between the two factors: moderate corruption of the note-event distribution does not impair transfer and can improve it, whereas broader renderer-side timbral support consistently improves out-of-domain generalization under a fixed note-event distribution. Finally, online-rendered examples complement real and existing synthetic data in a strong combined-data regime. These results suggest that synthetic AMT data should prioritize coverage of note-level attributes and their timbral realizations over realistic joint score structure.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv eess.AS, nuovi lavori del giorno).
Categorie arXiv: Audio and Speech Processing (eess.AS).
Pagina del lavoro: https://arxiv.org/abs/2610.11197 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-09T07:08:27.907Z dalla pagina https://arxiv.org/list/eess.AS/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).