Active inference and speech motor control

Area of Science:

• Neuroscience
• Computational Neuroscience
• Speech Science

Background:

• Active inference is a general brain theory minimizing sensory prediction errors.
• It has been widely applied to manual action control but less to speech motor control.
• Speech motor control is a complex sensorimotor function crucial for human interaction.

Purpose of the Study:

• To present the first detailed active inference framework for auditorily guided speech production.
• To compare active inference models with existing computational models of speech motor control.
• To explain adaptation and compensation mechanisms in speech following auditory feedback perturbations.

Main Methods:

• Developing an active inference framework for speech production.
• Comparing the architecture of active inference models to established speech motor control models.
• Modeling auditory feedback perturbations to observe adaptation and compensation.

Main Results:

• The study details an active inference framework for speech production.
• It provides an account of how speech adapts to auditory feedback changes.
• Active inference models offer unique insights into speech motor control mechanisms.

Conclusions:

• Active inference provides a unified framework for understanding speech motor control.
• It highlights the role of proprioception in speech motor learning.
• This approach can model phenomena like phonetic convergence and the influence of external voices.