Articulatory Kinematics

Articulatory kinematics studies the movement of speech organs during speech production, aiming to understand the relationship between these movements and the resulting acoustic signal. Current research focuses on developing sophisticated models, including neural network architectures like convolutive matrix factorization and generative adversarial networks (GANs), to analyze and represent articulatory data, often leveraging techniques like self-supervised learning. These advancements are improving our ability to model speech production, with implications for speech therapy, the design of assistive technologies, and a deeper understanding of the neural mechanisms underlying speech.