CoChat

Bias in Automated Speaker Recognition

Open access
Hutiri, Wiebke Toussaint, Ding, Aaron Yi · 2022 · 2022 ACM Conference on Fairness Accountability and Transparency
Cited by 47

The DOI matches its Crossref record; 12 of the 12 references it cites were found in a registry or an index.

Abstract

Automated speaker recognition uses data processing to identify speakers by their voice. Today, automated speaker recognition is deployed on billions of smart devices and in services such as call centres. Despite their wide-scale deployment and known sources of bias in related domains like face recognition and natural language processing, bias in automated speaker recognition has not been studied systematically. We present an in-depth empirical and analytical study of bias in the machine learning development workflow of speaker verification, a voice biometric and core task in automated speaker recognition. Drawing on an established framework for understanding sources of harm in machine learning, we show that bias exists at every development stage in the well-known VoxCeleb Speaker Recognition Challenge, including data generation, model building, and implementation. Most affected are female speakers and non-US nationalities, who experience significant performance degradation. Leveraging the insights from our findings, we make practical recommendations for mitigating bias in automated speaker recognition, and outline future research directions.

Overview

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

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

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

Integrity signals

Registry

Crossref · 10.1145/3531146.3533089

Cites

12 of 12 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

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

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

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

Self-check

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

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

Citations

Cites

Research works printed but not linked: 41

  1. Martine Adda-Decker and Lori Lamel. 2005. Do speech recognizers prefer female speakers?INTERSPEECH (2005), 2205–2208. https://www.isca-speech.org/archive/interspeech_2005/addadecker05_interspeech.html
  2. Tolga Bolukbasi, Kai-wei Chang, James Zou, Venkatesh Saligrama, and Adam Kalai. 2016. Man is to Computer Programmer as Woman is to Homemaker ? Debiasing Word Embeddings. In NIPS’16: Proceedings of the 30th International Conference on Neural Information Processing Systems. 4356 – 4364.
  3. Joy Buolamwini and Timnit Gebru. 2018. Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification. In Proceedings of Machine Learning Research: Conference on Fairness, Accountability, and Transparency, Vol. 81. 1889–1896.
  4. 10.21437/Interspeech.2020-1064 · DOI · Paper page
  5. Gianni Fenu, Mirko Marras, Giacomo Medda, and Giacomo Meloni. 2021. Fair Voice Biometrics : Impact of Demographic Imbalance on Group Fairness in Speaker Recognition. (2021), 1892–1896.
  6. Sadaoki Furui. 1994. An Overview of Speaker Recognition Technology. In ESCA Workshop on Automatic Speaker Recognition, Identification and Verification. 1 – 9.
  7. 10.1016/j.csl.2019.101032 · DOI · Paper page
  8. Moritz Hardt, Eric Price, and Nathan Srebro. 2016. Equality of opportunity in supervised learning. Advances in Neural Information Processing Systems (2016), 3323–3331.
  9. Khaled Hechmi Trung Ngo Trong Ville Hautamaki and Tomi Kinnunen. 2021. VoxCeleb Enrichment for Age and Gender Recognition. (2021). http://arxiv.org/abs/2109.13510 · DOI
  10. Georg Heigold, Ignacio Moreno, Samy Bengio, and Noam Shazeer. 2016. End-to-End Text-Dependent Speaker Verification. In ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 5115–5119.
  11. Sara Hooker Nyalleng Moorosi Gregory Clark Samy Bengio and Emily Denton. 2020. Characterising Bias in Compressed Models. https://arxiv.org/abs/2010.03058 · arXiv
  12. Julia Angwin, Jeff Larson, Surya Mattu and Lauren Kirchner. 2016. Machine Bias. https://www.propublica.org/article/machine-bias-risk-assessments-in-criminal-sentencing
  13. 10.1016/j.imavis.2013.10.001 · DOI
  14. 10.1109/ICB.2013.6613025 · DOI
  15. 10.5555/1577069.1755843 · doi:10.5555/1577069.1755843
  16. 10.1016/j.specom.2009.08.009 · DOI · Paper page
  17. Chao Li, Xiaokong Ma, Bing Jiang, Xiangang Li, Xuewei Zhang, Xiao Liu, Ying Cao, Ajay Kannan, and Zhenyao Zhu. 2017. Deep speaker: An end-to-end neural speaker embedding system. arXiv (2017).
  18. 10.21437/Interspeech.2020-2248 · DOI
  19. 10.3389/fbioe.2015.00126 · doi:10.3389/fbioe.2015.00126
  20. M McLaren L Ferrer D Castan and A Lawson. 2016. The Speakers in the Wild (SITW) speaker recognition database.. In Interspeech. pdfs.semanticscholar.org. https://pdfs.semanticscholar.org/3fe3/58a66359ee2660ec0d13e727eb8f3f0007c2.pdf
  21. Ninareh Mehrabi, Fred Morstatter, Nripsuta Saxena, Kristina Lerman, and Aram Galstyan. 2019. A survey on bias and fairness in machine learning. arXiv (2019).
  22. Arsha Nagrani Joon Son Chung Jaesung Huh Andrew Brown Ernesto Coto Weidi Xie Mitchell McLaren Douglas A Reynolds and Andrew Zisserman. 2020. VoxSRC 2020: The Second VoxCeleb Speaker Recognition Challenge. (2020). http://arxiv.org/abs/2012.06867 · arXiv
  23. 10.1016/j.csl.2019.101027 · DOI · Paper page
  24. Arsha Nagrani, Joon Son Chung, and Andrew Zisserman. 2017. Voxceleb: A large-scale speaker identification dataset. arXiv (2017), 2616–2620.
  25. 10.1016/j.csl.2019.06.001 · DOI
  26. 10.21437/Interspeech.2020-1815 · DOI
  27. NIST. 2019. NIST 2019 Speaker Recognition Evaluation Plan. 1 (2019), 1–7.
  28. NIST. 2020. NIST 2020 CTS Speaker Recognition Challenge Evaluation Plan. Technical Report. 1–8 pages.
  29. 10.21437/Interspeech.2016-523 · DOI
  30. Omkar M Parkhi, Andrea Vedaldi, and Andrew Zisserman. 2015. Deep Face Recognition. In British Machine Vision Conference.
  31. 10.21437/Interspeech.2020-1249 · DOI
  32. Inioluwa Deborah Raji and Genevieve Fried. 2021. About Face: A Survey of Facial Recognition Evaluation. (2021). http://arxiv.org/abs/2102.00813 · arXiv
  33. Douglas A. Reynolds. 2002. An Overview of Automatic Speaker Recognition Technology. IEEE (2002).
  34. Lea Schönherr, Maximilian Golla, Thorsten Eisenhofer, Jan Wiele, Dorothea Kolossa, and Thorsten Holz. 2020. Unacceptable, where is my privacy? Exploring Accidental Triggers of Smart Speakers. (8 2020). http://arxiv.org/abs/2008.00508 · arXiv
  35. Karen Simonyan and Andrew Zisserman. 2015. Very deep convolutional networks for large-scale image recognition. 3rd International Conference on Learning Representations, ICLR 2015 - Conference Track Proceedings(2015), 1–14.
  36. Rita Singh. 2019. Profiling Humans from their Voice. https://doi.org/10.1007/978-981-13-8403-5 · doi:10.1007/978-981-13-8403-5
  37. D Snyder, D Garcia-Romero, D Povey, and S Khudanpur. 2017. Deep Neural Network Embeddings for Text-Independent Speaker Verification.Interspeech (2017). https://www.isca-speech.org/archive/Interspeech_2017/pdfs/0620.PDF
  38. 10.1109/ICASSP.2018.8461375 · DOI · Paper page
  39. 10.21437/Interspeech.2017-1746 · DOI
  40. 10.1145/3485730.3493448 · DOI
  41. Wiebke Toussaint Akhil Mathur Fahim Kawsar and Aaron Yi Ding. 2022. Tiny always-on and fragile: Bias propagation through design choices in on-device machine learning workflows. (2022) 19 pages. http://arxiv.org/abs/2201.07677 · arXiv

Cited by

See all 47

Metadata

Authors
Hutiri, W. T., & Ding, A. Y.
Venue
2022 ACM Conference on Fairness Accountability and Transparency
Publisher
ACM
Published
2022-06-20
Pages
230-247
OpenAlex
W4226371168
Cited by
47
References
12
FWCI
10.79
Type
conference paper
Language
en
Open access
hybrid

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

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.