Correlates of vowel clarity in the spectrotemporal modulation domain: Application to speech impairment evaluation

Abstract

This article reports on vowel clarity metrics based on spectrotemporal modulations of speech signals. Motivated by previous findings on the relevance of modulation-based metrics for speech intelligibility assessment and pathology classification, the current study used factor analysis to identify regions within a bi-dimensional modulation space, the magnitude power spectrum, as in Elliott and Theunissen [(2009). PLoS Comput. Biol. 5(3), e1000302] by relating them to a set of conventional acoustic metrics of vowel space area and vowel distinctiveness. Two indices based on the energy ratio between high and low modulation rates across temporal and spectral dimensions of the modulation space emerged from the analyses. These indices served as input for measurements of central tendency and classification analyses that aimed to identify vowel-related speech impairments in French native speakers with head and neck cancer (HNC) and Parkinson dysarthria (PD). Following the analysis, vowel-related speech impairment was identified in HNC speakers, but not in PD. These results were consistent with findings based on subjective evaluations of speech intelligibility. The findings reported are consistent with previous studies indicating that impaired speech is associated with attenuation in energy in higher spectrotemporal modulation bands.

Aging of Amateur Singers and Non-singers: From Behavior to Resting-state Connectivity

“…The data detailed here represent a subset of a larger project. Other components of the project have been published elsewhere: a speech perception in noise task and structural imaging data (T1w and diffusion MRI data) (Perron et al, 2021;Perron et al, 2022) and a standardized passage reading task (Marczyk, Belley, et al, 2022;Marczyk, O'Brien, et al, 2022).…”

Do children laugh like their parents? Conversational laughter mimicry occurrence and acoustic alignment in middle-childhood

“…Spectral modulations characterise changes across frequency bands, having been shown to correspond with F0 ŕuctuations, harmonics, and formant patterns (Giraud and Poeppel, 2012;Rosen et al, 1992) with a perceptual upper limit around 3 cycles per octave (hereafter, c/o) (Dusan, 2007;Liu and Eddins, 2008) and a critical region of 0.8-1.3 c/o (Flinker et al, 2019). Numerous studies focused on unpacking the modulation representation of speech signals in the context of assessing speech intelligibility (Elhilali et al, 2003;Elliott and Theunissen, 2009;Flinker et al, 2019) and voice pathologies (Moro-Velázquez et al, 2015;Marczyk et al, 2022). Recent work by Ludusan and Wagner (2020b,a) showed that modulations of the amplitude envelope and F0 could be used to distinguish speech, laughter, and speech-laughter.…”

“…To obtain the MPS of laughs in our dataset, similar procedures described in Marczyk et al (2022) and were developed and applied. All processing was done in MATLAB 2021a (MathWorks Inc, USA) and based on adaptations to scripts described in Flinker et al (2019).…”

Differences between Mimicking and Non-Mimicking laughter in Child-Caregiver Conversation: A Distributional and Acoustic Analysis

“…Spectral modulations have been shown to correspond with F0 fluctuations, harmonics, and formant patterns (Giraud & Poeppel, 2012;Rosen, Carlyon, Darwin, & Russell, 1992) with perceptual limits around 3 cycles per octave (hereafter, c/o) (Dusan, 2007;Liu & Eddins, 2008) and a critical region of 0.8-1.3 c/o (Flinker et al, 2019). Numerous studies focused on unpacking the modulation representation of speech signals in the context of assessing speech intelligibility (Elhilali, Chi, & Shamma, 2003;Elliott & Theunissen, 2009;Flinker et al, 2019) and voice pathologies (Moro-Velázquez, Gómez-García, Godino Llorente, & Andrade-Miranda, 2015;Marczyk, O'Brien, Tremblay, Woisard, & Ghio, 2022). There have been but a handful of studies that use STM metrics to analyze anything other than speech.…”