A syllable-based analysis of speech temporal organization: a comparison between speaking styles in dysarthric and healthy populations

Abstract

A comparison of how healthy and dysarthric pathological speakers adapt their production is a way to better understand the processes and constraints that interact during speech production in general. The present study focuses on spontaneous speech obtained with varying recording scenarios from five different groups of speakers. Patients suffering from motor speech disorder (dysarthria) affecting speech production are compared to healthy speakers. Three types of dysarthria have been explored: Parkinson's Disease, Amyotrophic Lateral Sclerosis and Cerebellar ataxia. This paper first presents general figures based on syllable-level annotation mining, including detailed information about healthy/pathological speakers variability. Then, we report on the results of automatic timing parsing of interval sequences in speech syllable annotations performed using TGA (Time Group Analysis) methodology. We observed that mean syllable-based speaking rates in time groups for the healthy speakers were higher than those measured in the recordings of dysarthric speakers. The variability in timing patterns (duration regression slopes, intercepts, and nPVI) depended also on the speaking styles in particular populations.

A diachronic study on Italian speech rhythm in Parkinson’s Disease

“…As a first step towards this ambitious outcome of the research project, four speech samples were selected from the corpus, two recordings dated before the %V alteration (2010 and 2017) and two dated after it (2018 and 2021). In order to observe if and how the results obtained with an automatic procedure could diverge from those of the manual analysis, these speech samples were automatically annotated by means of the software SPPAS [26] -version 4.0, already used in researches on French dysarthric speech [27].…”

Accurate Synthesis of Dysarthric Speech for Asr Data Augmentation

“…To synthesize dysarthric speech, there is a need to build a system controlling different characteristics of dysarthric speech for generating variant dysarthric speech. As will be discussed later in this paper, and according to a number of studies (Rudzicz, Namasivayam et al 2012, Zhang, Dang et al 2014, Bigi, Klessa et al 2015, Kuo and Tjaden 2016, Yunusova, Graham et al 2016), such a system should have the following capabilities in order to support generation of authentic and diverse speech: 1) ability to control the speaking rate (duration), pitch, energy for a variety of dysarthria severity levels, 2) ability to learn and model pause behavior of dysarthric speakers (e.g., duration of pause and pause occurrence) and control pause insertion locations and durations 3) ability to learn and model individual voice characteristics of speakers and use these to generate new speaking styles 4) ability to learn and model these characteristics from a small amount of dysarthric speech data.…”