Vox-Profile: A Speech Foundation Model Benchmark for Characterizing Diverse Speaker and Speech Traits

TL;DR

Vox-Profile is a comprehensive benchmark for characterizing diverse speaker and speech traits using foundation models, enabling multi-dimensional profiling and supporting various speech analysis applications.

Contribution

It introduces a holistic, multi-dimensional speech trait benchmark grounded in speech science, developed with domain experts, and validated across multiple datasets and models.

Findings

Vox-Profile effectively characterizes static and dynamic speech traits.

It enhances analysis of ASR performance variability.

It evaluates speech generation systems with automated profiles.

Abstract

We introduce Vox-Profile, a comprehensive benchmark to characterize rich speaker and speech traits using speech foundation models. Unlike existing works that focus on a single dimension of speaker traits, Vox-Profile provides holistic and multi-dimensional profiles that reflect both static speaker traits (e.g., age, sex, accent) and dynamic speech properties (e.g., emotion, speech flow). This benchmark is grounded in speech science and linguistics, developed with domain experts to accurately index speaker and speech characteristics. We report benchmark experiments using over 15 publicly available speech datasets and several widely used speech foundation models that target various static and dynamic speaker and speech properties. In addition to benchmark experiments, we showcase several downstream applications supported by Vox-Profile. First, we show that Vox-Profile can augment existing speech recognition datasets to analyze ASR performance variability. Vox-Profile is also used as a tool to evaluate the performance of speech generation systems. Finally, we assess the quality of our automated profiles through comparison with human evaluation and show convergent validity. Vox-Profile is publicly available at: https://github.com/tiantiaf0627/vox-profile-release.