Abstract:Machine learning techniques have long been applied in many fields and have gained a lot of success. The purpose of learning processes is generally to obtain a set of parameters based on a given data set by minimizing a certain objective function which can explain the data set in a maximum likelihood or minimum estimation error sense. However, most of the learned parameters are highly data dependent and rarely reflect the true physical mechanism that is involved in the observation data. In order to obtain the inherent knowledge involved in the observed data, it is necessary to combine physical models with learning process rather than only fitting the observations with a black box model. To reveal underlying properties of human speech production, we proposed a learning process based on a physiological articulatory model and a coarticulation model, where both of the models are derived from human mechanisms. A two-layer learning framework was designed to learn the parameters concerned with physiological level using the physiological articulatory model and the parameters in the motor planning level using the coarticulation model. The learning process was carried out on an articulatory database of human speech production. The learned parameters were evaluated by numerical experiments and listening tests. The phonetic targets obtained in the planning stage provided an evidence for understanding the virtual targets of human speech production. As a result, the model based learning process reveals the inherent mechanism of the human speech via the learned parameters with certain physical meaning.

A simulation based parameter optimiza

A Simulation Based Parameter Optimization for a Coarticulation Model.

Observation and Modeling of Lingual Coarticulation in the Planning Stage

Decoding the dancing of the tongue: A model-based learning approach to phonetic targets in coarticulationa)

Decoding the Dancing of the Tongue: A Model-Based Learning Approach to Phonetic Targets in Coarticulation

A Model-Based Learning Process for Modeling Coarticulation of Human Speech

Implement of Coarticulation in Physiological Articulatory Model

Improved visual speech synthesis method of parameter control

Optimization Techniques for a Physical Model of Human Vocalisation

Investigation and modeling of coarticulation in speech production

Improvement of a Physiological Articulatory Model for Synthesis of Vowel Sequences

A Parametric Model for Voice Conversion

Improvements of a Physiological Articulatory Model in Construction and Control Strategy

Evaluating Features and Metrics for High-Quality Simulation of Early Vocal Learning of Vowels

A Design of Laryngeal Structures for a Physiological Articulatory Model

An Optimal Design Method of Articulator under Movement Trajectory Constraint

Target-filtering model based articulatory movement prediction for articulatory control of HMM-based speech synthesis

The Realization of Acoustic Model Based on Noise Modulating for Cochlear Implantation

Speech synthesis of VCV sequence using a physiological articulatory model

Vocal Tract Area Estimation by Gradient Descent

Articulatory Control of HMM-based Parametric Speech Synthesis Driven by Phonetic Knowledge