Robust and Explainable Depression Identification from Speech Using Vowel-Based Ensemble Learning Approaches

TL;DR

This paper presents vowel-based ensemble learning methods for depression detection from speech, emphasizing explainability and robustness, with approaches that decompose symptoms and severity for improved clinical utility.

Contribution

It introduces novel vowel-based embeddings and ensemble strategies that enhance explainability and robustness in depression classification from speech data.

Findings

Performance comparable to state-of-the-art baselines

Enhanced robustness against dataset mean/median variations

Improved system explainability for clinical use

Abstract

This study investigates explainable machine learning algorithms for identifying depression from speech. Grounded in evidence from speech production that depression affects motor control and vowel generation, pre-trained vowel-based embeddings, that integrate semantically meaningful linguistic units, are used. Following that, an ensemble learning approach decomposes the problem into constituent parts characterized by specific depression symptoms and severity levels. Two methods are explored: a "bottom-up" approach with 8 models predicting individual Patient Health Questionnaire-8 (PHQ-8) item scores, and a "top-down" approach using a Mixture of Experts (MoE) with a router module for assessing depression severity. Both methods depict performance comparable to state-of-the-art baselines, demonstrating robustness and reduced susceptibility to dataset mean/median values. System explainability benefits are discussed highlighting their potential to assist clinicians in depression diagnosis and screening.