Elamur Research⌕
☰
home / papers / 2026 / ToneSwiper: Facilitating Manual ToDI-annotation of Dutch Prosody

ToneSwiper: Facilitating Manual ToDI-annotation of Dutch Prosody

Summary

Expert manual intonation labeling is still needed to study how prosody is structured and what it means, and to train and evaluate automatic prosody labelers. Yet Dutch has no large public corpus annotated in the ToDI framework. The authors attribute this partly to the cost of working in general-purpose tools such as Praat and ELAN, where labeling relies heavily on the mouse and is slow. They introduce ToneSwiper, an open-source, cross-platform Python application built only for ToDI transcription. It combines audio playback, a spectrogram with pitch track, auto-scrolling and pitch-preserving slowdown. Annotators assign accents and boundaries by pressing arrow-key combinations in time with the audio. For example, up followed by down yields H*L. The key mapping can be reconfigured for ToBI-like systems in other languages, and the output imports into Praat as a point tier. A small pilot had two annotators label a hard 4-minute dialogue from the Dutch Map Task Corpus to assess speed and inter-rater agreement. Choosing labels while still listening proved difficult in such dense speech. The exact pilot figures were not in the excerpt provided.

Key points

  • ToneSwiper is distributed via the Python Package Index (version 0.3.3) and installs with pip. It targets categorical events (H/L accents, IP boundaries, downstep, delay), not paralinguistic pitch variation. Achievable precision is about ≤100 ms.
  • The authors cite three ergonomic gaps in Praat/ELAN: manual typing of labels, mouse-based placement, and no on-the-fly audio stretching.
  • Reported baseline costs:
    • ToBI labeling of American English takes 100–200× real time (Syrdal et al., 2001).
    • The authors estimate ToDI labeling in Praat at no less than 60×.
  • Automatic results for comparison:
    • Hu et al. (2020), Dutch ToDI: 94.6% accent detection and 75.4% classification, but only on scripted single-speaker data.
    • Rosenberg (2010), AutoBI on task-oriented dialogue: 73.5% detection and 69.8% classification.
  • The pilot used a 4-minute Dutch Map Task Corpus dialogue (Ladd and Schepman, 2003), which features rapid turn-taking, overlap and backchannels. The work is funded by NWO (project "Who's next?").

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 10.63317/5ctuktk6ffpi

@inproceedings{westera2026toneswiper,
  title = {{ToneSwiper: Facilitating Manual ToDI-annotation of Dutch Prosody}},
  author = {Westera, Matthijs and Reitsema, Ariëlle},
  booktitle = {Proceedings of the Fifteenth Language Resources and Evaluation Conference},
  year = {2026},
  doi = {10.63317/5ctuktk6ffpi},
  url = {https://doi.org/10.63317/5ctuktk6ffpi}
}

References

26 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Anne H. Anderson, Miles Bader, Ellen Gurman Bard, Elizabeth Boyle, Gwyneth Doherty, Simon Garrod, Stephen Isard, Jacqueline Kowtko, Jan McAllister, Jim Miller, et al. 1991. The HCRC map task corpus. Language and speech, 34(4):351– 366. Bastian Bechtold. Soundfile [Python package]. ↗
  2. Mary E. Beckman, Manuel Díaz-Campos, Julia Tevis McGory, and Terrell A. Morgan. 2002. Intonation across Spanish, in the Tones and Break Indices framework. ↗
  3. Paul Boersma and David Weenink. 2025. Praat: doing phonetics by computer. find
  4. David Brazil, Malcolm Coulthard, and Catherine Johns. 1980. Discourse intonation and language teaching. Longman, London. find
  5. Hendrik Buschmeier and Marcin Wlodarczak. 2013. TextGridTools: A TextGrid processing and analysis toolkit for Python. In Tagungsband der 24. Konferenz zur elektronischen sprachsignalverarbeitung (ESSV 2013). find
  6. Johanneke Caspers. 2003. Local speech melody as a limiting factor in the turn-taking system in dutch. Journal of Phonetics, 31(2):251–276. ↗
  7. Elisabeth Delais-Roussarie, Brechtje Post, Mathieu Avanzi, Carolin Buthke, Albert Di Cristo, Ingo Feldhausen, Sun-Ah Jun, Philippe Martin, Trudel Meisenburg, Annie Rialland, et al. 2015. Intonational phonology of French: Developing a ToBI system for French. Intonation in romance, pages 63–100. ↗
  8. ELAN. 2025. Version 7.0 [Computer software]. Max Planck Institute for Psycholinguistics, The Language Archive, Nijmegen. find
  9. Wendy Elvira-García, Paolo Roseano, Ana María Fernández-Planas, and Eugenio Martínez-Celdrán. 2016. A tool for automatic transcription of intonation: Eti_ToBI a ToBI transcriber for Spanish and Catalan. Language Resources and Evaluation, 50(4):767–792. ↗
  10. Carlos Gussenhoven. 1984. On the grammar and semantics of sentence accents, volume 16. De Gruyter Mouton, Berlin/New York. find
  11. Carlos Gussenhoven. 2005. Transcription of Dutch intonation. Prosodic typology: The phonology of intonation and phrasing, 118:145. Carlos Gussenhoven, Toni Rietveld, and Joop Kerkhoff. Transcription of Dutch Intonation, Second Edition, version 2.3. Paweł Głomski. PyLibRB [python package]. ↗
  12. Charles R. Harris, K. Jarrod Millman, Stéfan J. van der Walt, Ralf Gommers, Pauli Virtanen, David Cournapeau, Eric Wieser, Julian Taylor, Sebastian Berg, Nathaniel J. Smith, Robert Kern, Matti Picus, Stephan Hoyer, Marten H. van Kerkwijk, Matthew Brett, Allan Haldane, Jaime Fernández del Río, Mark Wiebe, Pearu Peterson, Pierre Gérard-Marchant, Kevin Sheppard, Tyler Reddy, Warren Weckesser, Hameer Abbasi, Christoph Gohlke, and Travis E. Oliphant. 2020. Array programming with NumPy. Nature, 585(7825):357–362. ↗ our card
  13. Xuliang He, Vincent J. van Heuven, and Carlos Gussenhoven. 2012. The selection of intonation contours by Chinese L2 speakers of Dutch: Orthographic closure vs. prosodic knowledge. Second Language Research, 28(3):283–318. ↗
  14. Na Hu, Berit Janssen, Jeff Hansen, Carlos Gussenhoven, and Aoju Chen. 2020. Automatic analysis of speech prosody in Dutch. In Interspeech, pages 155–159. find
  15. John D. Hunter. 2007. Matplotlib: A 2d graphics environment. Computing in Science & Engineering, 9(3):90–95. ↗
  16. Yannick Jadoul, Bill Thompson, and Bart de Boer. 2018. Introducing Parselmouth: A Python interface to Praat. Journal of Phonetics, 71:1–15. ↗
  17. D. R. Ladd and A. Schepman. 2003. Dutch map task corpus, 1999. SN: 4632. ↗
  18. Janet Pierrehumbert. 1980. The phonology and phonetics of English intonation. Thesis, Massachusetts Institute of Technology. Accepted: 2009-01-23T14:36:47Z. find
  19. Janet Pierrehumbert and Julia Hirschberg. 1990. The meaning of intonational contours in the interpretation of discourse. In Intentions in Communication, pages 271–311. MIT Press, Cambridge, MA. ↗
  20. John Pitrelli, Mary Beckman, and Julia Hirschberg. 1994. Evaluation of prosodic transcription labeling. In Proceeedings of the 3rd International Conference on Spoken Language Processing, pages 123–126. Yokohama, Japan. PyQt6. [Python package]. find
  21. Andrew Rosenberg. 2010. AutoBI: a tool for automatic ToBI annotation. In Interspeech, pages 146–149. Makuhari, Chiba, Japan. Rubber Band Library. [Computer software]. find
  22. Kim E.A. Silverman, Mary E. Beckman, John F. Pitrelli, Mari Ostendorf, Colin W. Wightman, Patti Price, Janet B. Pierrehumbert, and Julia Hirschberg. 1992. ToBI: a standard for labeling English prosody. In ICSLP, volume 2, pages 867–870. Sounddevice. [Python package]. find
  23. Ann K. Syrdal, Julia Hirschberg, Julie McGory, and Mary Beckman. 2001. Automatic ToBI prediction and alignment to speed manual labeling of prosody. Speech Communication, 33(1):135– 151. Speech Annotation and Corpus Tools. ↗
  24. Jennifer J. Venditti. 2005. The J_ToBI model of Japanese intonation. Prosodic typology: The phonology of intonation and phrasing, pages 172–200. ↗
  25. Matthijs Westera, Daniel Goodhue, and Carlos Gussenhhoven. 2020. Meanings of tones and tunes. In The Oxford Handbook of Language Prosody. Oxford University Press. ↗
  26. Wanyue Zhai and Mark Hasegawa-Johnson. 2023. Wav2ToBI: a new approach to automatic ToBI transcription. In Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH, volume 2023, pages 2748–2752. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/westera-2026-toneswiper-facilitating-manual-todi-annotation/