Elamur Research⌕
☰
home / papers / 2026 / Nos_Brais-GL: A FAIR Galician TTS Corpus for Neural Speech Synthesis

Nos_Brais-GL: A FAIR Galician TTS Corpus for Neural Speech Synthesis

Summary

Galician has reasonable text resources, but it lacks open, high-quality speech data for building neural TTS. Within Proxecto Nós, the authors release Nos_Brais-GL, a male single-speaker corpus. It complements the earlier female Nos_Celtia-GL and addresses that corpus's weaknesses: few exclamatory and imperative sentences, repetitive numeric and formulaic phrasing, and gaps in phonetic contexts. The voice talent was chosen through a listening test and confirmed with acoustic measurements. The script combines parliamentary transcripts, public TV scripts, dictionary examples, oral-archive transcriptions, an existing GTM/CRPIH dataset and literary excerpts, all revised to current Galician norms. It is deliberately smaller than its predecessor, because about 13,000 sentences had proven sufficient for good synthesis, which cut production time and cost. The paper also trains state-of-the-art TTS models and evaluates them subjectively and objectively, but the concrete scores are in the part of the text not provided here. Openly licensed and built on FAIR principles, the resource serves as a reusable model for TTS corpus design in minority languages.

Key points

  • Size: about 18 hours of studio recordings, with 16,121 utterances (168,320 words) read by one professional male speaker.
  • Speaker selection: 37 linguists and speech-technology experts rated three candidates using the FOLERPA tool. The chosen speaker had the best combined score, a speaking rate of about 2.8 words/s and a mean F0 of 172 Hz.
  • Domain mix (sentences):
    • Mixed: 6,075
    • GTM/CRPIH: 4,151
    • Literary: 1,500
    • TVG: 1,239
    • AGO: 1,224
    • Parlamento de Galicia: 780
    • DRAG: 652
    • Single words from the Dicionario de pronuncia da lingua galega: 500
  • Targeted enrichment: literary excerpts were added for under-represented phenomena, such as alveolar /n/ and varied imperative and exclamatory sentences.
  • Availability: distributed on Zenodo (DOI 10.5281/zenodo.14265241) under CC BY 4.0. It belongs to the Nós ecosystem alongside Nos_Celtia-GL and the Nós-TTS web interface.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.63317/2x6ou8cyanmo

@inproceedings{vladu2026nos,
  title = {{Nos\_Brais-GL: A FAIR Galician TTS Corpus for Neural Speech Synthesis}},
  author = {Vladu, Adina Ioana and Sánchez, Antonio Moscoso and Magariños, Carmen and Lago, María Perez and Rei, Elisa Fernández},
  booktitle = {Proceedings of the Fifteenth Language Resources and Evaluation Conference},
  year = {2026},
  doi = {10.63317/2x6ou8cyanmo},
  url = {https://doi.org/10.63317/2x6ou8cyanmo}
}

References

35 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Kaito Baba, Wataru Nakata, Yuki Saito, and Hiroshi Saruwatari. 2024. The t05 system for the voicemos challenge 2024: Transfer learning from deep image classifier to naturalness mos prediction of high-quality synthetic speech. 2024 IEEE Spoken Language Technology Workshop (SLT), pages 818–824. find
  2. Daniela Braga, Luís Coelho, Fernando Gil V. Resende, and Miguel Sales Dias. 2007a. Subjective and Objective Assessment of TTS Voice Font Quality. In Proc. find
  3. SPECOM 2007 – 12th International Conference Speech and Computer, pages 306–311, Moscow, Russia. find
  4. Daniela Braga, Luís Coelho, Fernando Gil V. Resende, and Miguel Sales Dias. 2007b. Subjective and Objective Evaluation of Brazilian Portuguese TTS Voice Font Quality. In Proc. find
  5. AST 2007 - 14th International Workshop on Advances in Speech Technology, pages 306–311, Maribor, Slovenia. find
  6. Erica Cooper and Junichi Yamagishi. 2021. How do voices from past speech synthesis challenges compare today? In 11th ISCA Speech Synthesis Workshop (SSW 11), pages 183–188. find
  7. CRPIH and GTM. 2023. CRPIH_UVigo-GL-Voices: Galician TTS dataset. https://doi.org/10.5281/zenodo.8027725. Dataset. Available under CC BY 4.0 license. — [doi:10.5281/zenodo.8027725](https://doi.org/10.5281/zenodo.8027725) ↗
  8. Iria de Dios-Flores, Carmen Magariños, Adina Ioana Vladu, John E. Ortega, José Ramom Pichel, Marcos García, Pablo Gamallo, Elisa Fernández Rei, Alberto Bugarín-Diz, Manuel González González, Senén Barro, and Xosé Luis Regueira. 2022. The Nós Project: Opening routes for the Galician language in the f… find
  9. Elisa Fernández Rei. 2016. Dialectal, Historical and Sociolinguistic Aspects of Galician Intonation. Dialectologia, 6:147–169. find
  10. Elisa Fernández Rei, Alba Aguete Cajiao, César Osorio Peláez, and Jose A. Cutrín Garabal. 2021. FOLERPA: Ferramenta On-Line para ExpeRimentación PerceptivA. https://ilg.usc.gal/folerpa. ↗
  11. Francisco Fernández Rei. 2025. Arquivo do galego oral. http://ilg.usc.es/ago. Director: Francisco Fernández Rei. Santiago de Compostela: Instituto da Lingua Galega. [Accessed: 2025]. ↗
  12. Noelia García Díaz, Marta Vázquez Abuín, Carmen Magariños, Adina Ioana Vladu, Antonio Moscoso Sánchez, and Elisa Fernández Rei. 2024. Nos_Celtia-GL: An Open High-Quality Speech Synthesis Resource for Galician. In Iber-SPEECH 2024, pages 91–95, ISCA. ISCA. find
  13. Manuel González González. 2008. O novo galego urbano. In Mercedes Brea, Francisco Fernández Rei, and Xosé Luís Regueira, editors, Cada palabra pesaba, cada palabra medía. Homenaxe a Antón Santamarina, pages 363–374. Universidade de Santiago de Compostela, Santiago de Compostela. find
  14. Manuel González González. 2025. Dicionario da Real Academia Galega. https://academia.gal/dicionario. Director: M. González González. A Coruña: Real Academia Galega. [Accessed: 2025]. ↗
  15. Wen-Chin Huang, Szu-Wei Fu, Erica Cooper, Ryandhimas E. Zezario, Tomoki Toda, Hsin-Min Wang, Junichi Yamagishi, and Yu Tsao. 2024. The voicemos challenge 2024: Beyond speech quality prediction. 2024 IEEE Spoken Language Technology Workshop (SLT), pages 803–810. find
  16. Jaehyeon Kim, Sungwon Kim, Jungil Kong, and Sungroh Yoon. 2020. Glow-TTS: A Generative Flow for Text-to-Speech via Monotonic Alignment Search. In Proc. of the 34th Int. Conf. on Neural Inf. Process. Syst., NIPS’20, Vancouver, Canada. Curran Associates Inc. find
  17. Jaehyeon Kim, Jungil Kong, and Juhee Son. Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speech. In Int. Conf. on Machine Learning. find
  18. Diederik P. Kingma and Max Welling. 2014. Autoencoding variational bayes. In 2nd International Conference on Learning Representations (ICLR), Banff, AB, Canada. Conference Track Proceedings. find
  19. J. Kominek and A. W. Black. 2003. The CMU ARCTIC Databases for Speech Synthesis Research. Technical report, Language Technologies Institute, Carnegie Mellon University, Pittsburgh, USA. find
  20. Jungil Kong, Jaehyeon Kim, and Jaekyoung Bae. 2020. HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis. NIPS’20. Curran Associates Inc. find
  21. Carmen Magariños, Alp Öktem, Antonio Moscoso Sánchez, Marta Vázquez Abuín, Noelia García Díaz, Adina Ioana Vladu, Elisa Fernández Rei, and María Baqueiro Vidal. 2024. Nós-TTS: a Web User Interface for Galician Text-to-Speech. In Proc. of the 16th Int. Conf. on Comput. Process. of Portuguese - Vol. 2… find
  22. Georgia Maniati, Alexandra Vioni, Nikolaos Ellinas, Karolos Nikitaras, Konstantinos Klapsas, June Sig Sung, Gunu Jho, Aimilios Chalamandaris, and Pirros Tsiakoulis. 2022. SOMOS: The samsung open mos dataset for the evaluation of neural text-to-speech synthesis. In Proc. find
  23. Interspeech 2022, pages 2388–2392. find
  24. Shivam Mehta, Ruibo Tu, Jonas Beskow, Éva Székely, and Gustav Eje Henter. 2024. Matcha-TTS: A fast TTS architecture with conditional flow matching. In Proc. ICASSP. find
  25. José Manuel Ramírez Sánchez and Carmen García-Mateo. 2022. Report on the Galician Language. Report D1.15, ELE. find
  26. Xosé Luís Regueira. 2009. Cambios fonéticos e fonolóxicos no galego contemporáneo. Estudos de Lingüística Galega, 1:147–167. find
  27. Xosé Luís Regueira. 2025. Dicionario de pronuncia da lingua galega. Director: Xosé Luís Regueira. Accessed 19/10/2025. find
  28. Rosalía Rodríguez Vázquez. 2019. La entonación de las preguntas parciales en una situación de contacto lingüístico: el caso del gallego y el español de galicia en hablantes bilingües. Estudios de fonética experimental, (28):81–124. find
  29. Eduardo Rodríguez Banga, Carmen García-Mateo, Francisco Méndez-Pazó, Manuel González-González, and Carmen Magariños. 2012. Cotovía: an open source TTS for Galician and Spanish. In VII Jornadas en Tecnología del Habla and III Iberian SLTech Workshop, IberSPEECH, pages 308–315. find
  30. A. Sesma and A. Moreno. 2000. CorpusCrt 1.0: Diseño de corpus orales equilibra-dos. http://gps-tsc.upc.es/veu/personal/sesma/CorpusCrt.php3. Computer program. ↗
  31. Xu Tan, Tao Qin, Frank K. Soong, and Tie-Yan Liu. 2021. A Survey on Neural Speech Synthesis. ArXiv, abs/2106.15561. — [arXiv:2106.15561](https://arxiv.org/abs/2106.15561) find
  32. A. I. Vladu, N. García Díaz, X. L. Regueira Fernández, C. Magariños, A. Moscoso Sánchez, D. Fernández López, E. Fernández Rei, and F. Dubert-García. 2025. Nos_Brais-GL. https://doi.org/10.5281/zenodo.14265241. Dataset. Available under CC BY 4.0 license. — [doi:10.5281/zenodo.14265241](https://doi.org/10.5281/zenodo.14265241) ↗
  33. Adina Ioana Vladu, Iria de Dios-Flores, Carmen Magariños, John E. Ortega, José Ramom Pichel, Marcos Garcia, Pablo Gamallo, Elisa Fernández Rei, Alberto Bugarín, Manuel González González, Senén Barro, and Xosé Luis Regueira. 2022. Proxecto Nós: Artificial intelligence at the service of the Galician l… find
  34. Marta Vázquez Abuín, Noelia García Díaz, Adina Ioana Vladu, Carmen Magariños, Adrián Vidal Miguéns, and Elisa Fernández Rei. 2023. Nos_Celtia-GL: Galician TTS corpus. https://doi.org/10.5281/zenodo.7716958. Dataset. Available under CC BY 4.0 license. — [doi:10.5281/zenodo.7716958](https://doi.org/10.5281/zenodo.7716958) ↗
  35. M. D. Wilkinson, M. Dumontier, I. J. Aalbersberg, G. Appleton, M. Axton, A. Baak, N. Blomberg, J.-W. Boiten, L. da Silva Santos, P. Bourne, et al. 2016. The FAIR Guiding Principles for Scientific Data Management and Stewardship. Scientific Data, 3:160018. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/vladu-2026-nos-brais-gl-fair-galician/