CoChat
Sign Up
Cited by 37Speech Recognition and SynthesisStuttering Research and TreatmentSpeech and dialogue systems

From User Perceptions to Technical Improvement: Enabling People Who Stutter to Better Use Speech Recognition

Open access
Lea, Colin, Huang, Zifang, Narain, Jaya, Tooley, Lauren, Yee, Dianna, Tran, Dung Tien, Georgiou, Panayiotis, Bigham, Jeffrey P +1 more·2023·Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems
DOI 10.1145/3544548.3581224
Open this paper in CoChat — free accountOpen PDF

Abstract

Consumer speech recognition systems do not work as well for many people with speech differences, such as stuttering, relative to the rest of the general population. However, what is not clear is the degree to which these systems do not work, how they can be improved, or how much people want to use them. In this paper, we first address these questions using results from a 61-person survey from people who stutter and find participants want to use speech recognition but are frequently cut off, misunderstood, or speech predictions do not represent intent. In a second study, where 91 people who stutter recorded voice assistant commands and dictation, we quantify how dysfluencies impede performance in a consumer-grade speech recognition system. Through three technical investigations, we demonstrate how many common errors can be prevented, resulting in a system that cuts utterances off 79.1% less often and improves word error rate from 25.4% to 9.9%.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1145/3544548.3581224

Cites

49 of 49 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • The present study was directly motivated by labeling issues identified in existing stuttered speech datasets.

    “The present study was directly motivated by labeling issues identified in existing stuttered speech datasets [57, 92].”

    From Disability-First AI Dataset Annotation: Co-designing Stuttered Speech Annotation Guidelines with People Who Stutter (2026) · cites this paper as [57]

    Section 5.4, “Limitations and Potential Biases” (page unavailable), identifies labeling errors in the paper’s own stuttered-speech data, including confused dysfluency types and discrepancies in intended-speech transcripts. It does not establish labeling issues across existing datasets. The citing authors’ statement about what motivated their present study is their own framing, not a finding this source can verify. (LLM, conf=0.78)

    Experimental — accuracy not yet measured

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

Research works printed but not linked: 25

  1. Sadeen Alharbi, Madina Hasan, Anthony JH Simons, Shelagh Brumftt, and Phil Green. 2018. A lightly supervised approach to detect stuttering in children's speech. In Interspeech 2018. ISCA, 3433-3437.
  2. Sebastian Bayerl, Florian Hönig, Joelle Reister, and Korbinian Riedhammer. 2020. Towards Automated Assessment of Stuttering and Stuttering Therapy. In Inter- national Conference on Text, Speech, and Dialogue.
  3. Sebastian P Bayerl, Alexander Wolf von Gudenberg, Florian Hönig, Elmar Nöth, and Korbinian Riedhammer. 2022. KSoF: The Kassel State of Fluency Dataset-A Therapy Centered Dataset of Stuttering. arXiv preprint arXiv:2203.05383 (2022). · arXiv
  4. Michael P Cannito, Annette Renee Burch, Christopher Watts, Patrick W Rappold, Stephen B Hood, and Kyla Sherrard. 1997. Disfuency in spasmodic dysphonia: A multivariate analysis. Journal of Speech, Language, and Hearing Research 40, 3 (1997), 627-641.
  5. William Chan, Navdeep Jaitly, Quoc V Le, and Oriol Vinyals. 2015. Listen, attend and spell. arXiv preprint arXiv:1508.01211 (2015). · arXiv
  6. Xi C Chen, Adithya Sagar, Justine T Kao, Tony Y Li, Christopher Klein, Stephen Pulman, Ashish Garg, and Jason D Williams. 2019. Active learning for domain classifcation in a commercial spoken personal assistant. Interspeech (2019).
  7. Frederic L Darley, Arnold E Aronson, and Joe R Brown. 1969. Diferential diag- nostic patterns of dysarthria. Journal of speech and hearing research 12, 2 (1969), 246-269.
  8. Arun Das, Jefrey Mock, Henry Chacon, Farzan Irani, Edward Golob, and Peyman Najafrad. 2020. Stuttering Speech Disfuency Prediction using Explainable Attribution Vectors of Facial Muscle Movements. In arXiv.
  9. Katie Deighton. 2021. Tech Firms Train Voice Assistants to Understand Atypical Speech. Wall Street Journal (online).
  10. Joseph R Dufy. 2019. Motor speech disorders e-book: Substrates, diferential diag- nosis, and management. Elsevier Health Sciences.
  11. Alexander M Goberman, Michael Blomgren, and Erika Metzger. 2010. Character- istics of speech disfuency in Parkinson disease. Journal of Neurolinguistics 23, 5 (2010), 470-478.
  12. Alex Graves. 2012. Sequence transduction with recurrent neural networks. arXiv preprint arXiv:1211.3711 (2012). · arXiv
  13. Kenneth Heafeld, Ivan Pouzyrevsky, Jonathan Clark, and Philipp Koehn. 2013. Scalable Modifed Kneser-Ney Language Model Estimation. In In Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pages 690-696. Association for Computational Linguistics.
  14. Peter Howell, Stephen Davis, and Jon Bartrip. 2009. The UCLASS archive of stuttered speech. Journal of Speech Language and Hearing Research 52 (2009), 556.
  15. Colin Lea, Vikramjit Mitra, Aparna Joshi, Sachin Kajarekar, and Jefrey P. Bigham. 2021. SEP-28k: A Dataset for Stuttering Event Detection From Podcasts With People Who Stutter. In ICASSP. IEEE.
  16. P. Mahesha and D.S. Vinod. 2016. Gaussian Mixture Model Based Classifcation of Stuttering Dysfuencies. Journal of Intelligent Systems 25, 3 (2016), 387-399.
  17. Roisin McNaney, Christopher Bull, Lynne Mackie, Floriane Dahman, Helen Stringer, Dan Richardson, and Daniel Welsh. 2018. StammerApp: Designing a Mobile Application to Support Self-Refection and Goal Setting for People Who Stammer. In Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems. ACM, 1-12.
  18. Nan Bernstein Ratnera and Brian MacWhinney. 2018. Fluency Bank: A new resource for fuency research and practice. Journal of Fluency Disorders (2018).
  19. Glyndon D Riley. 2009. SSI-4 stuttering severity instrument fourth edition.
  20. Abi Roper, Stephanie Wilson, Timothy Neate, and Jane Marshall. 2019. Speech and Language. In Web Accessibility. Springer, 121-131.
  21. Olabanji Shonibare, Xiaosu Tong, and Venkatesh Ravichandran. 2022. Enhancing ASR for Stuttered Speech with Limited Data Using Detect and Pass. arXiv preprint arXiv:2202.05396 (2022). · arXiv
  22. Jimmy Tobin and Katrin Tomanek. 2022. Personalized Automatic Speech Recog- nition Trained on Small Disordered Speech Datasets. In ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 6637-6641.
  23. Wayne Xiong, Jasha Droppo, Xuedong Huang, Frank Seide, Mike Seltzer, An- dreas Stolcke, Dong Yu, and Geofrey Zweig. 2016. Achieving Human Parity in Conversational Speech Recognition. CoRR abs/1610.05256. arXiv:1610.05256 · arXiv
  24. W. Xiong, L. Wu, F. Alleva, J. Droppo, X. Huang, and A. Stolcke. 2018. The Microsoft 2017 Conversational Speech Recognition System. In IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE.
  25. Cheng Yi, Shiyu Zhou, and Bo Xu. 2021. Efciently fusing pretrained acoustic and linguistic encoders for low-resource speech recognition. In IEEE Signal Processing Letters, Vol. 28. IEEE, 788-792.

Cited by

Sorted by citations

Related

Authors
Lea, C., Huang, Z., Narain, J., Tooley, L., Yee, D., Tran, D. T., Georgiou, P., Bigham, J. P., & Findlater, L.
Affiliations
  • Apple (United States) (US)
  • University of Washington (US)
ORCID
Venue
Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems
Publisher
ACM
Published
2023-04-19
Pages
1-16
DOI
10.1145/3544548.3581224
OpenAlex
W4366547420
Cited by
35
References
49
Citations by year
2026: 9 · 2025: 10 · 2024: 14 · 2023: 2
FWCI
6.20
Type
conference paper
Language
en
Open access
Licence
www.acm.org/publications/policies/copyright_policy#Background (vor)
Locations
  • CHI Conference on Human Factors in Computing Systems (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • DSpace@MIT (Massachusetts Institute of Technology) (submittedVersion, cc-by-nc)
  • arXiv (Cornell University)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Speech Recognition and Synthesis
  • Social Sciences › Psychology › Clinical Psychology › Stuttering Research and Treatment
  • Physical Sciences › Computer Science › Artificial Intelligence › Speech and dialogue systems
Keywords
stuttering, speech recognition, disfluency, voice assistants, user perception, word error rate, speech prediction, assistive technology

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review