Toward Fairness in Speech Recognition: Discovery and mitigation of performance disparities
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2022
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866909097916366848 |
|---|---|
| author | Dheram, Pranav Ramakrishnan, Murugesan Raju, Anirudh Chen, I-Fan King, Brian Powell, Katherine Saboowala, Melissa Shetty, Karan Stolcke, Andreas |
| author_facet | Dheram, Pranav Ramakrishnan, Murugesan Raju, Anirudh Chen, I-Fan King, Brian Powell, Katherine Saboowala, Melissa Shetty, Karan Stolcke, Andreas |
| contents | As for other forms of AI, speech recognition has recently been examined with respect to performance disparities across different user cohorts. One approach to achieve fairness in speech recognition is to (1) identify speaker cohorts that suffer from subpar performance and (2) apply fairness mitigation measures targeting the cohorts discovered. In this paper, we report on initial findings with both discovery and mitigation of performance disparities using data from a product-scale AI assistant speech recognition system. We compare cohort discovery based on geographic and demographic information to a more scalable method that groups speakers without human labels, using speaker embedding technology. For fairness mitigation, we find that oversampling of underrepresented cohorts, as well as modeling speaker cohort membership by additional input variables, reduces the gap between top- and bottom-performing cohorts, without deteriorating overall recognition accuracy. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2207_11345 |
| institution | arXiv |
| publishDate | 2022 |
| record_format | arxiv |
| spellingShingle | Toward Fairness in Speech Recognition: Discovery and mitigation of performance disparities Dheram, Pranav Ramakrishnan, Murugesan Raju, Anirudh Chen, I-Fan King, Brian Powell, Katherine Saboowala, Melissa Shetty, Karan Stolcke, Andreas Computation and Language Sound Audio and Speech Processing As for other forms of AI, speech recognition has recently been examined with respect to performance disparities across different user cohorts. One approach to achieve fairness in speech recognition is to (1) identify speaker cohorts that suffer from subpar performance and (2) apply fairness mitigation measures targeting the cohorts discovered. In this paper, we report on initial findings with both discovery and mitigation of performance disparities using data from a product-scale AI assistant speech recognition system. We compare cohort discovery based on geographic and demographic information to a more scalable method that groups speakers without human labels, using speaker embedding technology. For fairness mitigation, we find that oversampling of underrepresented cohorts, as well as modeling speaker cohort membership by additional input variables, reduces the gap between top- and bottom-performing cohorts, without deteriorating overall recognition accuracy. |
| title | Toward Fairness in Speech Recognition: Discovery and mitigation of performance disparities |
| topic | Computation and Language Sound Audio and Speech Processing |
| url | https://arxiv.org/abs/2207.11345 |