Acoustic impacts of geometric approximation at the level of velum and epiglottis on french vowels

In this work we study the effect of the velum and epiglottis on speech production of five French vowels. Our purpose is to examine whether it is possible to simplify the geometry of the vocal tract in the framework of articulatory synthesis to achieve a simpler geometric description without changing the acoustic properties. In the present study, we use MRI to acquire the 3D shape of the vocal tract with simultaneous recording of the speech signal. The geometric two-dimensional shape derived from these data was used as an input of numerical acoustic simulations. The geometrical shape was edited at the level of epiglottis and velum (with or without epiglottis, with or without a constant wall approximation at velum) and the spectra obtained via numerical acoustic simulations were compared with those obtained from audio recordings. This allows the impact of these articulators and geometrical simplifications to be assessed.

[1]  Xabier Jaureguiberry,et al.  The Flexible Audio Source Separation Toolbox Version 2.0 , 2014, ICASSP 2014.

[2]  Paolo Cignoni,et al.  MeshLab: an Open-Source Mesh Processing Tool , 2008, Eurographics Italian Chapter Conference.

[3]  B. Cox,et al.  Modeling power law absorption and dispersion for acoustic propagation using the fractional Laplacian. , 2010, The Journal of the Acoustical Society of America.

[4]  A. Rendell,et al.  Use of multiple GPUs on shared memory multiprocessors for ultrasound propagation simulations , 2012 .

[5]  Alan L. Yuille,et al.  Region Competition: Unifying Snakes, Region Growing, and Bayes/MDL for Multiband Image Segmentation , 1996, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  Guido Gerig,et al.  User-guided 3D active contour segmentation of anatomical structures: Significantly improved efficiency and reliability , 2006, NeuroImage.

[7]  Bradley E. Treeby,et al.  Full-wave nonlinear ultrasound simulation in an axisymmetric coordinate system using the discrete sine and cosine transforms , 2013, 2013 IEEE International Ultrasonics Symposium (IUS).

[8]  Yves Laprie,et al.  Comparison between 2D and 3D models for speech production: a study of french vowels , 2019 .

[9]  Guillermo Sapiro,et al.  Geodesic Active Contours , 1995, International Journal of Computer Vision.

[10]  Olov Engwall,et al.  Influence of vocal tract geometry simplifications on the numerical simulation of vowel sounds. , 2016, The Journal of the Acoustical Society of America.

[11]  Milan Sonka,et al.  3D Slicer as an image computing platform for the Quantitative Imaging Network. , 2012, Magnetic resonance imaging.

[12]  B T Cox,et al.  k-Wave: MATLAB toolbox for the simulation and reconstruction of photoacoustic wave fields. , 2010, Journal of biomedical optics.

[13]  Peter Birkholz,et al.  A three-dimensional model of the vocal tract for speech synthesis , 2003 .

[14]  Yves Laprie,et al.  Extension of the single-matrix formulation of the vocal tract: Consideration of bilateral channels and connection of self-oscillating models of the vocal folds with a glottal chink , 2016, Speech Commun..

[15]  Bradley E Treeby Acoustic attenuation compensation in photoacoustic tomography using time-variant filtering , 2013, Journal of biomedical optics.

[16]  Christine Ericsdotter DETAIL IN VOWEL AREA FUNCTIONS , 2007 .

[17]  Emmanuel Vincent,et al.  A General Flexible Framework for the Handling of Prior Information in Audio Source Separation , 2012, IEEE Transactions on Audio, Speech, and Language Processing.

[18]  Shinji Maeda,et al.  Compensatory Articulation During Speech: Evidence from the Analysis and Synthesis of Vocal-Tract Shapes Using an Articulatory Model , 1990 .