Quantifying Bias in Automatic Speech Recognition

TL;DR

This paper systematically quantifies biases in a Dutch state-of-the-art ASR system across gender, age, and accents, revealing specific error patterns and proposing mitigation strategies.

Contribution

It introduces a comprehensive bias analysis framework for ASR systems, focusing on phoneme-level errors and articulation differences, which is novel in the context of Dutch ASR.

Findings

Bias varies significantly across gender, age, and accents.

Phoneme-level errors reveal articulation-related biases.

Bias mitigation strategies are proposed based on findings.

Abstract

Automatic speech recognition (ASR) systems promise to deliver objective interpretation of human speech. Practice and recent evidence suggests that the state-of-the-art (SotA) ASRs struggle with the large variation in speech due to e.g., gender, age, speech impairment, race, and accents. Many factors can cause the bias of an ASR system. Our overarching goal is to uncover bias in ASR systems to work towards proactive bias mitigation in ASR. This paper is a first step towards this goal and systematically quantifies the bias of a Dutch SotA ASR system against gender, age, regional accents and non-native accents. Word error rates are compared, and an in-depth phoneme-level error analysis is conducted to understand where bias is occurring. We primarily focus on bias due to articulation differences in the dataset. Based on our findings, we suggest bias mitigation strategies for ASR development.