The Balancing Act: Unmasking and Alleviating ASR Biases in Portuguese

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kulkarni, Ajinkya, Tokareva, Anna, Qureshi, Rameez, Couceiro, Miguel
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917587728728064
author Kulkarni, Ajinkya
Tokareva, Anna
Qureshi, Rameez
Couceiro, Miguel
author_facet Kulkarni, Ajinkya
Tokareva, Anna
Qureshi, Rameez
Couceiro, Miguel
contents In the field of spoken language understanding, systems like Whisper and Multilingual Massive Speech (MMS) have shown state-of-the-art performances. This study is dedicated to a comprehensive exploration of the Whisper and MMS systems, with a focus on assessing biases in automatic speech recognition (ASR) inherent to casual conversation speech specific to the Portuguese language. Our investigation encompasses various categories, including gender, age, skin tone color, and geo-location. Alongside traditional ASR evaluation metrics such as Word Error Rate (WER), we have incorporated p-value statistical significance for gender bias analysis. Furthermore, we extensively examine the impact of data distribution and empirically show that oversampling techniques alleviate such stereotypical biases. This research represents a pioneering effort in quantifying biases in the Portuguese language context through the application of MMS and Whisper, contributing to a better understanding of ASR systems' performance in multilingual settings.
format Preprint
id arxiv_https___arxiv_org_abs_2402_07513
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle The Balancing Act: Unmasking and Alleviating ASR Biases in Portuguese
Kulkarni, Ajinkya
Tokareva, Anna
Qureshi, Rameez
Couceiro, Miguel
Computation and Language
Artificial Intelligence
Computers and Society
In the field of spoken language understanding, systems like Whisper and Multilingual Massive Speech (MMS) have shown state-of-the-art performances. This study is dedicated to a comprehensive exploration of the Whisper and MMS systems, with a focus on assessing biases in automatic speech recognition (ASR) inherent to casual conversation speech specific to the Portuguese language. Our investigation encompasses various categories, including gender, age, skin tone color, and geo-location. Alongside traditional ASR evaluation metrics such as Word Error Rate (WER), we have incorporated p-value statistical significance for gender bias analysis. Furthermore, we extensively examine the impact of data distribution and empirically show that oversampling techniques alleviate such stereotypical biases. This research represents a pioneering effort in quantifying biases in the Portuguese language context through the application of MMS and Whisper, contributing to a better understanding of ASR systems' performance in multilingual settings.
title The Balancing Act: Unmasking and Alleviating ASR Biases in Portuguese
topic Computation and Language
Artificial Intelligence
Computers and Society
url https://arxiv.org/abs/2402.07513