Metraipolis › archive › dc_bddc4eaeca67

TUTTI: Toward generalizable audio-to-score transcription via fully synthesized data — arXiv:2609.00640

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.00640
License (as archived): CC0-1.0
Archived: 2026-10-17T06:33:34.516Z
SHA-256 of the archived text: a8b38ae0e7b2caf0e8305bcfd13b51983b32f49cf8565b994f7e6e4b735c1606
Stable URL of this page: https://metraipolis.com/documento/dc_bddc4eaeca67

How to cite

TUTTI: Toward generalizable audio-to-score transcription via fully synthesized data — arXiv:2609.00640 — arXiv (en), https://arxiv.org/abs/2609.00640 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_bddc4eaeca67 (Metraipolis archive dc_bddc4eaeca67, sha256 a8b38ae0e7b2caf0e8305bcfd13b51983b32f49cf8565b994f7e6e4b735c1606)

BibTeX

@misc{dcbddc4eaeca67,
  title = {TUTTI: Toward generalizable audio-to-score transcription via fully synthesized data — arXiv:2609.00640},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.00640},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_bddc4eaeca67; sha256 a8b38ae0e7b2caf0e8305bcfd13b51983b32f49cf8565b994f7e6e4b735c1606},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.00640. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Generalizable Audio-to-Score (A2S) transcription is fundamentally constrained by the severe scarcity of high-quality, real-world paired data. Relying solely on existing human-annotated datasets often restricts the generalization of A2S models, limiting their efficacy primarily to single-instrumentation domains. To break this dependency on scarce real-world data, we introduce TUTTI (Transformer for Unified audio-To-score Transcription trained on Synthetic multi-Instrumentation Data), a pre-training paradigm driven by a purely synthetic, large-scale dataset. Rather than using human-composed scores, we leverage a symbolic music generation model to generate a massive, highly scalable multi-instrumentation corpus and create audio-score pairs with expressive acoustic characteristics. Capitalizing on the generated data, we employ a standard Transformer encoder-decoder architecture. We empirically demonstrate that pre-training a unified attention-based model on generated, multi-instrumentation data yields a consistently stronger foundational representation than single-instrumentation training. When fine-tuned with downstream real-world datasets, TUTTI outperforms previous approaches, establishing new overall state-of-the-art results across various A2S baselines. Notably, TUTTI shows remarkable cross-instrument transferability, effectively adapting to unseen instruments with highly competitive performance. The source code and the TuttiCorpus dataset will be made publicly available at this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-02 in cs.AI).
Categorie arXiv: Sound (cs.SD) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.00640 — 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 2026-10-17T06:15:12.810Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).