The TYPALOC Corpus: A Collection of Various Dysarthric Speech Recordings in Read and Spontaneous Styles

The TYPALOC Corpus: A Collection of Various Dysarthric Speech Recordings in Read and Spontaneous Styles Meunier C. (1), Fougeron C. (2), Fredouille C. (3), Bigi B. (1), Crevier-Buchman L. (2), Delais-Roussarie E. (4), Georgeton L. (1), Ghio A. (1), Laaridh I. (1,3), Legou T. (1), Pillot-Loiseau C. (2), Pouchoulin G. (1,3) (1) Aix Marseille Université, CNRS, LPL UMR 7309, 13100, Aix-en-Provence, France (2) Laboratoire de Phonétique et Phonologie, UMR 7018 CNRS-Université Paris 3 (3) University of Avignon, CERI/LIA, France (4) Laboratoire de Linguistique Formelle, UMR 7110 CNRS – Université Paris E-mail: {christine.meunier, brigitte.bigi, alain.ghio, thierry.legou, gilles.pouchoulin}@lpl-aix.fr, {cecile.fougeron, lise.buchman, claire.pillot}@univ-paris3.fr, {corinne.fredouille, laaridh.imed}@univ-avignon.fr, elisabeth.roussarie@wanadoo.fr, laurianne.georgeton@gmail.com Abstract This paper presents the TYPALOC corpus of French Dysarthric and Healthy speech and the rationale underlying its constitution. The objective is to compare phonetic variation in the speech of dysarthric vs. healthy speakers in different speech conditions (read and unprepared speech). More precisely, we aim to compare the extent, types and location of phonetic variation within these different populations and speech conditions. The TYPALOC corpus is constituted of a selection of 28 dysarthric patients (three different pathologies) and of 12 healthy control speakers recorded while reading the same text and in a more natural continuous speech condition. Each audio signal has been segmented into Inter-Pausal Units. Then, the corpus has been manually transcribed and automatically aligned. The alignment has been corrected by an expert phonetician. Moreover, the corpus benefits from an automatic syllabification and an Automatic Detection of Acoustic Phone-Based Anomalies. Finally, in order to interpret phonetic variations due to pathologies, a perceptual evaluation of each patient has been conducted. Quantitative data are provided at the end of the paper. Keywords: Corpus, speech disorders, read and spontaneous speech, automatic speech processing, anomaly detection 1. Introduction This paper presents the TYPALOC corpus of French Dysarthric and Healthy speech and the rationale underlying its constitution. This corpus is built in the context of the TYPALOC project (ANR-12-BSHS2-0003) whose objective is to compare phonetic variation in the speech of dysarthric vs. healthy speakers in different speech conditions. Assuming that pathological alterations in the speech signal are somewhat comparable in shape (under-articulated forms) to the one we can find in casual speech produced by healthy speakers, our objective is to compare the extent, types and location of phonetic variation within these different populations and speech conditions. Dysarthria refers to a group of motor speech disorders caused by damage to either the central or peripheral nervous system. These lesions disrupt the transfer of information from the nervous system to the muscles involved in speech production. According to the location of the lesion in the brain and the severity of the damage, any of the different speech subsystems can be affected (respiration, phonation, subglottal activities) and speech distortions can cover a large spectrum of degree. Moreover, according to the etiology of the lesion and the disease associated with it (toxic, metabolic, traumatic or degenerative diseases), types of dysarthria vary with respect to pathophysiologies determining the kind of deficits in the motor execution …

Content truncated to 500 words