Metraipolis › archive › dc_97a8bcd3dc3f
Tone on a Budget: A Reference-Free Metric for Lexical Tone in Massively Multilingual Text-to-Speech — arXiv:2609.14817
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.14817
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-15T23:15:31.316Z
- SHA-256 of the archived text
0fcb964e40c6f256fc882007dfd2dbea7521c85016eb703a4abbd878a9b19a59- Stable URL of this page
https://metraipolis.com/documento/dc_97a8bcd3dc3f
How to cite
Tone on a Budget: A Reference-Free Metric for Lexical Tone in Massively Multilingual Text-to-Speech — arXiv:2609.14817 — arXiv (en), https://arxiv.org/abs/2609.14817 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_97a8bcd3dc3f (Metraipolis archive dc_97a8bcd3dc3f, sha256 0fcb964e40c6f256fc882007dfd2dbea7521c85016eb703a4abbd878a9b19a59)
BibTeX
@misc{dc97a8bcd3dc3f,
title = {Tone on a Budget: A Reference-Free Metric for Lexical Tone in Massively Multilingual Text-to-Speech — arXiv:2609.14817},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.14817},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_97a8bcd3dc3f; sha256 0fcb964e40c6f256fc882007dfd2dbea7521c85016eb703a4abbd878a9b19a59},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.14817. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
In Yorùbá, pitch alone separates \d{o}k\d{o} (husband, Mid), \d{o}k\d{ò} (vehicle, Low), and \d{o}k\d{ó} (hoe, High) -- the diacritics ARE the tone marks. Yet character error rate (CER), the standard automated metric for text-to-speech (TTS), is in practice computed from ASR output that drops those marks: a synthesizer can ace CER and still say vehicle for husband. We introduce DunDun -- named for the dùndún, the Yorùbá talking drum that speaks through pitch alone -- an automated, reference-free lexical-tone metric that needs no tone-labelled corpus. The gold High/Mid/Low sequence is read from the input text's diacritics (in TTS that text exists by construction, so no reference recording is needed); the prediction comes from the audio's pitch track. We validate three ways. Flattening pitch with PSOLA resynthesis collapses DunDun while CER does not move. Inverting High and Low in the answer key of 300 native recordings drives the two-class readout to 0.14, symmetrically below its 0.35 chance level -- a consistency check on the scoring path, not independent evidence. And three native listeners, over 67 blind A/B trials, pick the tone-correct clip 89.6% of the time (95% CI 80.0-94.8; p < 1e-4); whether DunDun tracks those judgements trial by trial is not resolved at this sample size. Applied to a massively multilingual zero-shot TTS model, DunDun shows what CER cannot: Yorùbá tone sits near the native anchor before any Yorùbá fine-tuning (0.567 +/- 0.02 over five decode seeds vs. 0.596; chance 0.33), despite the 21.4% CER the model's own paper reports; and a few hours of clean audio halve CER (5.6% to 2.7% by 5h, 1.7% by 15h) while tone saturates within the hour. On non-tonal Swahili, CER already captures the gains: the metric a language needs is language-dependent. We release the metric and the complete validation protocol.
Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-15 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.14817 — 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 2026-10-15T22:51:09.908Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-15?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).