Temporal envelope and fine structure cues for dysarthric speech detection using CNNs

TL;DR

This paper introduces a novel approach for dysarthric speech detection by separately analyzing temporal envelope and fine structure cues inspired by human auditory processing, leading to improved detection accuracy.

Contribution

It proposes a new method that factors speech signals into envelope and fine structure for CNN-based dysarthric speech detection, outperforming traditional magnitude spectrum methods.

Findings

Processing both envelope and fine structure improves detection accuracy.

Separately learned representations outperform single-cue approaches.

The method aligns with human auditory perception mechanisms.

Abstract

Deep learning-based techniques for automatic dysarthric speech detection have recently attracted interest in the research community. State-of-the-art techniques typically learn neurotypical and dysarthric discriminative representations by processing time-frequency input representations such as the magnitude spectrum of the short-time Fourier transform (STFT). Although these techniques are expected to leverage perceptual dysarthric cues, representations such as the magnitude spectrum of the STFT do not necessarily convey perceptual aspects of complex sounds. Inspired by the temporal processing mechanisms of the human auditory system, in this paper we factor signals into the product of a slowly varying envelope and a rapidly varying fine structure. Separately exploiting the different perceptual cues present in the envelope (i.e., phonetic information, stress, and voicing) and fine structure (i.e., pitch, vowel quality, and breathiness), two discriminative representations are learned through a convolutional neural network and used for automatic dysarthric speech detection. Experimental results show that processing both the envelope and fine structure representations yields a considerably better dysarthric speech detection performance than processing only the envelope, fine structure, or magnitude spectrum of the STFT representation.