Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Gender Fairness in Audio Deepfake Detection: Performance and Disparity Analysis

Aishwarya Fursule, Shruti Kshirsagar, Anderson R. Avila

TL;DR

Results show that even when the overall EER difference between genders appears low, fairness-aware evaluation reveals disparities in error distribution that are obscured by aggregate performance measures, demonstrating that reliance on standard metrics is unreliable, whereas fairness metrics provide critical insights into demographic-specific failure modes.

Abstract

Audio deepfake detection aims to detect real human voices from those generated by Artificial Intelligence (AI) and has emerged as a significant problem in the field of voice biometrics systems. With the ever-improving quality of synthetic voice, the probability of such a voice being exploited for illicit practices like identity thest and impersonation increases. Although significant progress has been made in the field of Audio Deepfake Detection in recent times, the issue of gender bias remains underexplored and in its nascent stage In this paper, we have attempted a thorough analysis of gender dependent performance and fairness in audio deepfake detection models. We have used the ASVspoof 5 dataset and train a ResNet-18 classifier and evaluate detection performance across four different audio features, and compared the performance with baseline AASIST model. Beyond conventional metrics such as Equal Error Rate (EER %), we incorporated five established fairness metrics to quantify gender disparities in the model. Our results show that even when the overall EER difference between genders appears low, fairness-aware evaluation reveals disparities in error distribution that are obscured by aggregate performance measures. These findings demonstrate that reliance on standard metrics is unreliable, whereas fairness metrics provide critical insights into demographic-specific failure modes. This work highlights the importance of fairness-aware evaluation for developing a more equitable, robust, and trustworthy audio deepfake detection system.

Gender Fairness in Audio Deepfake Detection: Performance and Disparity Analysis

TL;DR

Results show that even when the overall EER difference between genders appears low, fairness-aware evaluation reveals disparities in error distribution that are obscured by aggregate performance measures, demonstrating that reliance on standard metrics is unreliable, whereas fairness metrics provide critical insights into demographic-specific failure modes.

Abstract

Audio deepfake detection aims to detect real human voices from those generated by Artificial Intelligence (AI) and has emerged as a significant problem in the field of voice biometrics systems. With the ever-improving quality of synthetic voice, the probability of such a voice being exploited for illicit practices like identity thest and impersonation increases. Although significant progress has been made in the field of Audio Deepfake Detection in recent times, the issue of gender bias remains underexplored and in its nascent stage In this paper, we have attempted a thorough analysis of gender dependent performance and fairness in audio deepfake detection models. We have used the ASVspoof 5 dataset and train a ResNet-18 classifier and evaluate detection performance across four different audio features, and compared the performance with baseline AASIST model. Beyond conventional metrics such as Equal Error Rate (EER %), we incorporated five established fairness metrics to quantify gender disparities in the model. Our results show that even when the overall EER difference between genders appears low, fairness-aware evaluation reveals disparities in error distribution that are obscured by aggregate performance measures. These findings demonstrate that reliance on standard metrics is unreliable, whereas fairness metrics provide critical insights into demographic-specific failure modes. This work highlights the importance of fairness-aware evaluation for developing a more equitable, robust, and trustworthy audio deepfake detection system.
Paper Structure (15 sections, 6 equations, 3 figures, 6 tables)

This paper contains 15 sections, 6 equations, 3 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Methodology
  4. Implication of Fairness Metrics in Gender-aware Audio Deepfake Detection
  5. Fairness Metrics
  6. Experimental Setup
  7. Dataset
  8. Feature Extraction
  9. Classifier and Training Protocol
  10. Evaluation Protocol
  11. Statistical Analysis
  12. RESULT AND DISCUSSION
  13. Fairness Analysis Across Gender
  14. Performance Differences Across Gender
  15. CONCLUSION

Figures (3)

  • Figure 1: The five group fairness metrics adopted in this study.
  • Figure 2: Balanced gender distribution for the ASVspoof5 dataset across train, development, and evaluation splits.
  • Figure 3: Illustration of the procedure adopted to perform group fairness assessment. Model performance is based on 5 group fairness measures.