mlsanityData to 7 Oct 2026

"OK Aura, Be Fair With Me": Demographics-Agnostic Training for Bias Mitigation in Wake-up Word Detection

arXiv 2604.05830 ↗First posted 7 Apr 2026cs.CL cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Voice-based interfaces are widely used; however, achieving fair Wake-up Word detection across diverse speaker populations remains a critical challenge due to persistent demographic biases. This study evaluates the effectiveness of demographics-agnostic training techniques in mitigating performance disparities among speakers of varying sex, age, and accent. We utilize the OK Aura database for our experiments, employing a training methodology that excludes demographic labels, which are reserved for evaluation purposes. We explore (i) data augmentation techniques to enhance model generalization and (ii) knowledge distillation of pre-trained foundational speech models. The experimental results indicate that these demographics-agnostic training techniques markedly reduce demographic bias, leading to a more equitable performance profile across different speaker groups. Specifically, one of the evaluated techniques achieves a Predictive Disparity reduction of 39.94% for sex, 83.65% for age, and 40.48% for accent when compared to the baseline. This study highlights the effectiveness of label-agnostic methodologies in fostering fairness in Wake-up Word detection.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 29 of its 52 references that are in the map. First posted on arXiv 7 Apr 2026.

Builds on

29 of 52 references.

References · 29

  1. 2022Robust Speech Recognition via Large-Scale Weak Supervisioncited by 3,972
  2. 2019SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognitioncited by 1,355
  3. 2022Self-Supervised Speech Representation Learning: A Reviewcited by 258
  4. 2017Word Embeddings Quantify 100 Years of Gender and Ethnic Stereotypescited by 251
  5. 2023Seamless: Multilingual Expressive and Streaming Speech Translationcited by 190
  6. 2022Comparative layer-wise analysis of self-supervised speech modelscited by 155
  7. 2024ASVspoof 5: Crowdsourced Speech Data, Deepfakes, and Adversarial Attacks at Scalecited by 88
  8. 2021Large-scale Self-Supervised Speech Representation Learning for Automatic Speaker Verificationcited by 71
  9. 2021Quantifying Bias in Automatic Speech Recognitioncited by 50
  10. 2021Deep Spoken Keyword Spotting: An Overviewcited by 43
  11. 2021SpecMix : A Mixed Sample Data Augmentation method for Training withTime-Frequency Domain Featurescited by 30
  12. 2022Toward Fairness in Speech Recognition: Discovery and mitigation of performance disparitiescited by 27
17 more
  1. 2024Towards measuring fairness in speech recognition: Fair-Speech datasetcited by 26
  2. 2024Twists, Humps, and Pebbles: Multilingual Speech Recognition Models Exhibit Gender Performance Gapscited by 24
  3. 2020Gender Bias in Depression Detection Using Audio Featurescited by 21
  4. 2024Leveraging Self-Supervised Learning for Speaker Diarizationcited by 21
  5. 2021Don't speak too fast: The impact of data bias on self-supervised speech modelscited by 20
  6. 2024On the social bias of speech self-supervised modelscited by 16
  7. 2024Lost in Transcription: Identifying and Quantifying the Accuracy Biases of Automatic Speech Recognition Systems Against Disfluent Speechcited by 15
  8. 2023Exploring data augmentation in bias mitigation against non-native-accented speechcited by 11
  9. 2023Device-Robust Acoustic Scene Classification via Impulse Response Augmentationcited by 10
  10. 2024A Contrastive Learning Approach to Mitigate Bias in Speech Modelscited by 9
  11. 2021FilterAugment: An Acoustic Environmental Data Augmentation Methodcited by 8
  12. 2022Tiny, always-on and fragile: Bias propagation through design choices in on-device machine learning workflowscited by 8
  13. 2025Fairness of Automatic Speech Recognition: Looking Through a Philosophical Lenscited by 5
  14. 2022Iterative pseudo-forced alignment by acoustic CTC loss for self-supervised ASR domain adaptationcited by 4
  15. 2023A Brief Tutorial on Sample Size Calculations for Fairness Auditscited by 4
  16. 2023Robust Wake-Up Word Detection by Two-stage Multi-resolution Ensemblescited by 1
  17. 2024On the Use of Audio to Improve Dialogue Policiescited by 1

Built on by

Not cited yet.