AequeVox: Automated Fairness Testing of Speech Recognition Systems

Abstract

AbstractAutomatic Speech Recognition (ASR) systems have become ubiquitous. They can be found in a variety of form factors and are increasingly important in our daily lives. As such, ensuring that these systems are equitable to different subgroups of the population is crucial. In this paper, we introduce,AequeVox, an automated testing framework for evaluating the fairness of ASR systems.AequeVoxsimulates different environments to assess the effectiveness of ASR systems for different populations. In addition, we investigate whether the chosen simulations are comprehensible to humans. We further propose a fault localization technique capable of identifying words that are not robust to these varying environments. Both components ofAequeVoxare able to operate in the absence of ground truth data.We evaluateAequeVoxon speech from four different datasets using three different commercial ASRs. Our experiments reveal that non-native English, female and Nigerian English speakers generate109%,528.5%and156.9%more errors, on average than native English, male and UK Midlands speakers, respectively. Our user study also reveals that 82.9% of the simulations (employed through speech transformations) had a comprehensibility rating above seven (out of ten), with the lowest rating being 6.78. This further validates the fairness violations discovered byAequeVox. Finally, we show that the non-robust words, as predicted by the fault localization technique embodied inAequeVox, show223.8%more errors than the predicted robust words across all ASRs.