Document Open Access Logo

Evaluating the Prosodic Diversity of TTS Models for L2 Prosody Assessment

Authors Mariana Julião , Alberto Abad , Helena Moniz

PDF

File

Thumbnail PDF
PDF
OASIcs.SLATE.2026.2.pdf
  • Filesize: 1.07 MB
  • 13 pages

Document Identifiers

Subject Classification

ACM Subject Classification
  • Applied computing → Computer-assisted instruction
  • Applied computing → Education
Keywords
  • Speech synthesis
  • TTS
  • prosody
  • L1
  • L2

Metrics

Abstract

Recent advances in text-to-speech (TTS) have led to synthetic speech that is often indistinguishable from natural speech at the level of individual utterances. However, it remains unclear whether such systems reproduce the prosodic variability observed in natural speech in a large native-speaker population. This question is particularly relevant for applications in computer-assisted language learning (CALL), as variability is a central property of prosody and a prerequisite for robust assessment.
In this work, we investigate whether TTS can approximate the distribution of prosodic patterns found in native speech, and whether it can serve as a reference for evaluating second language (L2) productions. To this end, we first compare different TTS models to native speakers, and then compare L2 speakers to synthetic speech by the model previously seen as the closest to native speakers.
Our results show that regardless of TTS achieving high perceptual quality, its prosodic variability substantially differs from that of native speakers. As a consequence, comparisons between L2 speech and TTS-based references reveal both the potential and the limitations of using synthetic speech for prosody assessment.

Cite As Get BibTex

Mariana Julião, Alberto Abad, and Helena Moniz. Evaluating the Prosodic Diversity of TTS Models for L2 Prosody Assessment. In 15th Symposium on Languages, Applications and Technologies (SLATE 2026). Open Access Series in Informatics (OASIcs), Volume 144, pp. 2:1-2:13, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2026) https://doi.org/10.4230/OASIcs.SLATE.2026.2

Author Details

Mariana Julião
  • HLT, INESC-ID, Lisbon, Portugal
Alberto Abad
  • HLT, INESC-ID, Lisbon, Portugal
  • Instituto Superior Técnico, Lisbon, Portugal
Helena Moniz
  • HLT, INESC-ID, Lisbon, Portugal
  • Faculdade de Letras da Universidade de Lisboa, Portugal
  • CLUL, Lisbon, Portugal

Funding

Work supported by Portuguese national funds through Fundação para a Ciência e a Tecnologia, I.P. (FCT) under projects UID/50021/2025 (DOI: https://doi.org/10.54499/UID/50021/2025) and UID/PRR/50021/2025 (DOI: https://doi.org/10.54499/UID/PRR/50021/2025), by CLUL UID/214/2025 (DOI: https://doi.org/10.54499/UID/00214/2025) and by the Portuguese Recovery and Resilience Plan and NextGenerationEU European Union funds under project C644865762-00000008 (ACCELERAT.AI).

References

  1. Tiago Bione, Jennica Grimshaw, and Walcir Cardoso. An evaluation of TTS as a pedagogical tool for pronunciation instruction: the ‘foreign’language context. CALL in a climate of change: adapting to turbulent global conditions-short papers from EUROCALL, pages 56-61, 2017. Google Scholar
  2. Jeisson Alonso Rodríguez Bonces. Implementing text-to-speech technology as a means of enhancing L2 reading fluency. EDU REVIEW. International Education and Learning Review/Revista Internacional de Educación y Aprendizaje, 1(2):61-73, 2019. Google Scholar
  3. Edresson Casanova, Kelly Davis, Eren Gölge, Görkem Göknar, Iulian Gulea, Logan Hart, Aya Aljafari, Joshua Meyer, Reuben Morais, Samuel Olayemi, et al. XTTS: a Massively Multilingual Zero-Shot Text-to-Speech Model. arXiv preprint arXiv:2406.04904, 2024. URL: https://arxiv.org/abs/2406.04904.
  4. Zhihao Du, Qian Chen, Shiliang Zhang, Kai Hu, Heng Lu, Yexin Yang, Hangrui Hu, Siqi Zheng, Yue Gu, Ziyang Ma, et al. Cosyvoice: A scalable multilingual zero-shot text-to-speech synthesizer based on supervised semantic tokens. arXiv preprint arXiv:2407.05407, 2024. URL: https://arxiv.org/abs/2407.05407.
  5. Zhihao Du, Changfeng Gao, Yuxuan Wang, Fan Yu, Tianyu Zhao, Hao Wang, Xiang Lv, Hui Wang, Xian Shi, Keyu An, et al. CosyVoice 3: Towards In-the-wild Speech Generation via Scaling-up and Post-training. arXiv preprint arXiv:2505.17589, 2025. URL: https://arxiv.org/abs/2505.17589.
  6. Wendy Elvira-García, Simone Balocco, Paolo Roseano, and Ana Ma Fernandez-Planas. ProDis: A dialectometric tool for acoustic prosodic data. Speech Communication, 97:9-18, 2018. URL: https://doi.org/10.1016/J.SPECOM.2017.12.013.
  7. Florian Eyben, Klaus R Scherer, Björn W Schuller, Johan Sundberg, Elisabeth André, Carlos Busso, Laurence Y Devillers, Julien Epps, Petri Laukka, Shrikanth S Narayanan, et al. The Geneva minimalistic acoustic parameter set (GeMAPS) for voice research and affective computing. IEEE transactions on affective computing, 7(2):190-202, 2015. URL: https://doi.org/10.1109/TAFFC.2015.2457417.
  8. Zöe Handley. Is Text-to-Speech synthesis ready for use in Computer-Assisted Language Learning? Speech Communication, 51(10):906-919, 2009. URL: https://doi.org/10.1016/J.SPECOM.2008.12.004.
  9. Florian Hönig, Anton Batliner, and Elmar Nöth. Automatic assessment of nonnative prosody-annotation, modelling and evaluation. In International Symposium on Automatic Detection of Errors in Pronunciation Training (IS ADEPT), pages 21-30, 2012. Google Scholar
  10. Florian Hönig, Anton Batliner, Karl Weilhammer, and Elmar Nöth. Islands of Failure: Employing word accent information for pronunciation quality assessment of English L2 learners. In International Workshop on Speech and Language Technology in Education, 2009. Google Scholar
  11. Na Hu, Jiseung Kim, Riccardo Orrico, Stella Gryllia, and Amalia Arvaniti. Can OpenAI’s TTS model convey information status using intonation like humans? In Speech Prosody 2024, pages 32-36, 2024. URL: https://doi.org/10.21437/SpeechProsody.2024-7.
  12. Kiyoshi Ichihara, James C Boyd, IFCC Committee on Reference Intervals, and Decision Limits (C-RIDL). An appraisal of statistical procedures used in derivation of reference intervals. Clinical chemistry and laboratory medicine, 48(11):1537-1551, 2010. Google Scholar
  13. Daniel Korzekwa, Jaime Lorenzo-Trueba, Thomas Drugman, and Bozena Kostek. Computer-assisted pronunciation training — Speech synthesis is almost all you need. Speech Communication, 142:22-33, 2022. URL: https://doi.org/10.1016/J.SPECOM.2022.06.003.
  14. Denis Liakin, Walcir Cardoso, and Natallia Liakina. The pedagogical use of mobile speech synthesis (TTS): Focus on French liaison. Computer Assisted Language Learning, 30(3-4):325-342, 2017. Google Scholar
  15. Michael McAuliffe, Michaela Socolof, Sarah Mihuc, Michael Wagner, and Morgan Sonderegger. Montreal Forced Aligner: Trainable Text-Speech Alignment Using Kaldi. In Proc. Interspeech 2017, pages 498-502, 2017. URL: https://doi.org/10.21437/Interspeech.2017-1386.
  16. Wolfgang Menzel, Eric Atwell, Patrizia Bonaventura, Daniel Herron, Peter Howarth, Rachel Morton, and Clive Souter. The ISLE corpus of non-native spoken English. In Proceedings of LREC 2000: Language Resources and Evaluation Conference, vol. 2, pages 957-964. European Language Resources Association, 2000. Google Scholar
  17. Björn Schuller, Stefan Steidl, Anton Batliner, Simone Hantke, Florian Hönig, J. R. Orozco-Arroyave, Elmar Nöth, Yue Zhang, and Felix Weninger. The INTERSPEECH 2015 computational paralinguistics challenge: nativeness, Parkinson’s & eating condition. In Interspeech 2015, pages 478-482, 2015. URL: https://doi.org/10.21437/Interspeech.2015-179.
  18. Cristian Tejedor-Garcıa, David Escudero-Mancebo, César González-Ferreras, Enrique Cámara-Arenas, and Valentín Cardeñoso-Payo. Evaluating the efficiency of synthetic voice for providing corrective feedback in a pronunciation training tool based on minimal pairs. Proc. SLaTE, Stockholm, Sweden, pages 26-30, 2017. Google Scholar
  19. Steven Weinberger. Speech Accent Archive. Retrieved from http://accent.gmu.edu., 2015.
  20. Yujia Xiao and Frank K Soong. Proficiency Assessment of ESL Learner’s Sentence Prosody with TTS Synthesized Voice as Reference. In INTERSPEECH, pages 1755-1759, 2017. URL: https://doi.org/10.21437/INTERSPEECH.2017-64.
  21. Junichi Yamagishi, Christophe Veaux, and Kirsten MacDonald. CSTR VCTK Corpus: English Multi-speaker Corpus for CSTR Voice Cloning Toolkit (version 0.92), 2019. [sound]. URL: https://doi.org/10.7488/ds/2645.
  22. Stephen A Zahorian and Hongbing Hu. A spectral/temporal method for robust fundamental frequency tracking. The Journal of the Acoustical Society of America, 123(6):4559-4571, 2008. Google Scholar
  23. Yixuan Zhou, Guoyang Zeng, Xin Liu, Xiang Li, Renjie Yu, Ziyang Wang, Runchuan Ye, Weiyue Sun, Jiancheng Gui, Kehan Li, Zhiyong Wu, and Zhiyuan Liu. VoxCPM: Tokenizer-Free TTS for Context-Aware Speech Generation and True-to-Life Voice Cloning. arXiv preprint arXiv:2509.24650, 2025. URL: https://doi.org/10.48550/arXiv.2509.24650.
Any Issues?
X

Feedback on the Current Page

CAPTCHA

Thanks for your feedback!

Feedback submitted to Dagstuhl Publishing

Could not send message

Please try again later or send an E-mail
© 2023-2026 Schloss Dagstuhl – LZI GmbH Schloss Dagstuhl – LZI GmbH About DROPS Imprint Privacy Contact