The Balancing Act: Unmasking and Alleviating ASR Biases in Portuguese - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2024

The Balancing Act: Unmasking and Alleviating ASR Biases in Portuguese

Anna Tokareva
  • Fonction : Auteur
  • PersonId : 1346491
Mohammed Rameez Qureshi
  • Fonction : Auteur
  • PersonId : 1344061
Miguel Couceiro

Résumé

In the field of spoken language understanding, systems like Whisper and Multilingual Massive Speech (MMS) have shown state-of-theart 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.
Fichier principal
Vignette du fichier
Final_Camera_Ready_LT_EDI_2024_ASR_BIAS (1) (1).pdf (1.19 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04436147 , version 1 (05-02-2024)

Licence

Identifiants

  • HAL Id : hal-04436147 , version 1

Citer

Ajinkya Kulkarni, Anna Tokareva, Mohammed Rameez Qureshi, Miguel Couceiro. The Balancing Act: Unmasking and Alleviating ASR Biases in Portuguese. EACL 2024 LT-EDI WorkShop, Mar 2024, St. Julians, Malta. ⟨hal-04436147⟩
65 Consultations
67 Téléchargements

Partager

Gmail Mastodon Facebook X LinkedIn More