Efficient Long Speech Sequence Modelling for Time-Domain Depression Level Estimation

TL;DR

This paper introduces an efficient time-domain approach for long speech sequence modeling to estimate depression severity, avoiding information loss from traditional time-frequency methods and capturing long-range cues in raw audio signals.

Contribution

It proposes a novel time-domain deep learning framework with state space, dual-path, and external attention modules for depression level estimation from long speech signals.

Findings

Outperforms state-of-the-art methods on AVEC2013 and AVEC2014 datasets.

Effectively captures long-range depression cues in raw audio.

Demonstrates robustness in modeling real-world speech patterns.

Abstract

Depression significantly affects emotions, thoughts, and daily activities. Recent research indicates that speech signals contain vital cues about depression, sparking interest in audio-based deep-learning methods for estimating its severity. However, most methods rely on time-frequency representations of speech which have recently been criticized for their limitations due to the loss of information when performing time-frequency projections, e.g. Fourier transform, and Mel-scale transformation. Furthermore, segmenting real-world speech into brief intervals risks losing critical interconnections between recordings. Additionally, such an approach may not adequately reflect real-world scenarios, as individuals with depression often pause and slow down in their conversations and interactions. Building on these observations, we present an efficient method for depression level estimation using long speech signals in the time domain. The proposed method leverages a state space model coupled with the dual-path structure-based long sequence modelling module and temporal external attention module to reconstruct and enhance the detection of depression-related cues hidden in the raw audio waveforms. Experimental results on the AVEC2013 and AVEC2014 datasets show promising results in capturing consequential long-sequence depression cues and demonstrate outstanding performance over the state-of-the-art.