Meta-Learning Approaches for Speaker-Dependent Voice Fatigue Models

TL;DR

This paper explores meta-learning techniques to improve speaker-dependent voice fatigue models, demonstrating that advanced methods like transformers outperform traditional models in health monitoring tasks.

Contribution

It introduces a reformulation of speaker adaptation as a meta-learning problem and compares three novel approaches using speech embeddings.

Findings

All meta-learning methods outperformed traditional models.

Transformer-based approach achieved the best performance.

Meta-learning improves speaker-dependent fatigue prediction accuracy.

Abstract

Speaker-dependent modelling can substantially improve performance in speech-based health monitoring applications. While mixed-effect models are commonly used for such speaker adaptation, they require computationally expensive retraining for each new observation, making them impractical in a production environment. We reformulate this task as a meta-learning problem and explore three approaches of increasing complexity: ensemble-based distance models, prototypical networks, and transformer-based sequence models. Using pre-trained speech embeddings, we evaluate these methods on a large longitudinal dataset of shift workers (N=1,185, 10,286 recordings), predicting time since sleep from speech as a function of fatigue, a symptom commonly associated with ill-health. Our results demonstrate that all meta-learning approaches tested outperformed both cross-sectional and conventional mixed-effects models, with a transformer-based method achieving the strongest performance.