Title

Speech modeling and processing. Control of a Speech Robot via an Optimum Neural–Network–Based Internal Model with Constraints. Statistical tools in speech processing.

Abstract

This Ph.D. dissertation deals with speech modeling and processing, which both share the speech quality aspect. An optimum internal model with constraints is proposed and discussed for the control of a biomechanical speech robot based on the equilibrium point hypothesis (EPH, λ–model). It is supposed that the robot internal space is composed of the motor commands λ of the equilibrium point hypothesis. The main idea of the work is that the robot movements, and in particular the robot speech production, are carried out in such a way that, the length of the path traveled in the internal space, is minimized, under acoustical and mechanical constraints. Mathematical aspect of the problem leads to one of the problems of variational calculus, the so–called geodesic problem, whose exact analytical solution is quite complicated. By using some empirical findings, an approximate solution for the proposed optimum internal model is then developed and implemented. It gives interesting and challenging results, and shows that the proposed internal model is quite realistic; namely, some similarities are found between the robot speech and the real one. Next, by aiming to analyze speech signal, several methods of statistical speech signal processing are developed. They are based on higher–order statistics (namely, on normalized central moments and on the fourth–order cumulant), as well as on discrete normalized entropy. In this framework, we also designed an unbiased and efficient estimator of the fourth–order cumulant in both batch and adaptive versions.

Keywords

Robotics, optimal and adaptive control, speech production, speech motor control, optimum task planning, artificial intelligence, equilibrium point hypothesis (EPH, λ–model), artificial neural networks (ANN), internal model, mathematical physics, variational calculus, constrained optimization, geodesic problem, Euler–Lagrange equation, differential equations, statistical signal processing, estimation, adaptive estimation, bias, MSE, higher–order statistics, cumulants, moments, entropy.

Acknowledgment

Acknowledgment for the help and other things goes to the following persons (see electronic version of the Ph.D. dissertation for details): Eric Moreau, Salah Bourennane, Yannick Deville, Jean–Christophe Pesquet, Michel Paindavoine, Sylvain Maire, Nicolas Maubec, Béatrice Marulier, Marina Schelkounova (Марина Щелкунова), Noëlle Maitre, Moustafa Belqasmi, Alexei N. Popov (Алексей Н. Попов), Dimitri Andreevich Lissatchenko (Дмитрий Андреевич Лисаченко), Sergei Nikolaevich Manida (Сергей Николаевич Манида), François Brut, Nicolas André, Bleicke Holm, Guillaume Gibert, Romain Marret, Jacques Schneider, Bernard Xerri, Jean Barrère, Bruno Borloz, Christian Tavernier, Nadège Thirion–Moreau, Paola Goatin, Rémi Dubroca, Virginie Attina, Antoine Serrurier, Jana Brunner, Luis Hernández Gómez, Raphaëlle Lirou, Caroline Dufy, Estelle Lezean, and especially, to my grand–farthers.

Download

You can download my Ph.D. dissertation from the official site Theses.fr of the French agence bibliographique de l'enseignement supérieur (ABES). You can also find my Ph.D. dissertation in the French catalog of higher education and research libraries (Sudoc).