What problem does this paper attempt to address?

The problem that this paper attempts to solve is: how to simulate the production of VCV (vowel - consonant - vowel) sequences through an inverse biomechanical model. Specifically, the author focuses on how to transform the perceptual targets of speech signals into specific motor commands during the speech generation process, and also considers the influence of the physical characteristics of the articulatory organs on this process. ### Main problems 1. **Diversity in speech generation**: - For a given phoneme, its physical characteristics will vary according to different speech environments, speaking styles or speech rates. - In order to generate specific acoustic features, theoretically, multiple articulatory configurations can be selected. 2. **Limitations of existing research**: - Existing research usually needs to specify the desired trajectories of the entire time series, which is more complex in practical applications. - Existing research fails to fully consider the influence of the inherent physical characteristics of the articulatory organs on the trajectories in the perceptual space. ### Solutions To solve these problems, the author proposes a goal - based control model, which consists of three main parts: 1. **Static forward model**: - It describes the relationship from motor commands to acoustic characteristics. By using a two - dimensional biomechanical tongue model, it simulates the influence of tongue muscle activities on acoustic features. 2. **Target definition in the perceptual space**: - It defines the target areas of each phoneme in the perceptual space, and these areas contain all perceptually equivalent acoustic realizations. 3. **Planning procedure based on optimization principles**: - It uses optimization techniques to infer an appropriate sequence of motor commands from a phoneme sequence. By minimizing a cost function \( C \), it ensures that the changes in motor commands are as smooth as possible and can produce the expected perceptual effects. ### Formula representation - The cost function \( C \) is defined as: \[ C = C_{\text{speaker}}+C_{\text{listener}} \] where: - \( C_{\text{speaker}} \) is the speaker - oriented cost, representing the global path length of connecting the target commands of each phoneme, which is used to keep the changes in motor commands minimized. - \( C_{\text{listener}} \) is the listener - oriented cost, ensuring the expected perceptual effects of each phoneme. In this way, the author provides a framework that can quantitatively test various hypotheses about speech coarticulation and evaluate the effects of different models.

Significant enhancement of room temperature ferromagnetism in surfactant coated polycrystalline Mn doped ZnO particles

A 3D biomechanical vocal tract model to study speech production control: How to take into account the gravity?

Modéliser le physique pour comprendre le contrôle : le cas de l'anticipation en production de parole

A 3D dynamical biomechanical tongue model to study speech motor control

Modélisation biomécanique tri-dimensionnelle de l'articulateur lingual pour étudier le contrôle de la parole

Biomechanical modeling of the human tongue

An articulatory model of standard Chinese using MRI and X-ray movie

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

Controlling chaotic vocal fold oscillations in the numerical production of vowel sounds

Complete reconstruction of the tongue contour through acoustic to articulatory inversion using real-time MRI data

Acoustic VR in the Mouth: A Real-Time Speech-Driven Visual Tongue System.

Use of a biomechanical tongue model to predict the impact of tongue surgery on speech production

Laryngeal Muscular Control of Vocal Fold Posturing: Numerical Modeling and Experimental Validation.

Progress in animation of an EMA-controlled tongue model for acoustic-visual speech synthesis

A Speech-Driven 3-D Tongue Model with Realistic Movement in Mandarin Chinese.

Inverse Biomechanical Modeling of the Tongue via Machine Learning and Synthetic Training Data

Emergent Jaw Predominance in Vocal Development through Stochastic Optimization

A Multilinear Tongue Model Derived from Speech Related MRI Data of the Human Vocal Tract

On the Emergence of Phonological Knowledge and on Motor Planning and Motor Programming in a Developmental Model of Speech Production

Simulating Articulatory Trajectories with Phonological Feature Interpolation

A biomechanical modeling study of the effects of the orbicularis oris muscle and jaw posture on lip shape