Gender Bias in Instruction-Guided Speech Synthesis Models

TL;DR

This paper investigates gender bias in instruction-guided speech synthesis models, revealing that models tend to reinforce stereotypes in interpreting occupation-related prompts, with bias levels varying by model size.

Contribution

The study provides empirical evidence of gender bias in TTS models' interpretation of occupation prompts and compares bias across different model sizes.

Findings

Models exhibit gender bias in occupation prompts

Bias varies with model size

Bias tends to reinforce stereotypes

Abstract

Recent advancements in controllable expressive speech synthesis, especially in text-to-speech (TTS) models, have allowed for the generation of speech with specific styles guided by textual descriptions, known as style prompts. While this development enhances the flexibility and naturalness of synthesized speech, there remains a significant gap in understanding how these models handle vague or abstract style prompts. This study investigates the potential gender bias in how models interpret occupation-related prompts, specifically examining their responses to instructions like "Act like a nurse". We explore whether these models exhibit tendencies to amplify gender stereotypes when interpreting such prompts. Our experimental results reveal the model's tendency to exhibit gender bias for certain occupations. Moreover, models of different sizes show varying degrees of this bias across these occupations.