The FACTS model of speech motor control: Fusing state estimation and task-based control

Area of Science:

• Computational Neuroscience
• Speech Science
• Robotics

Background:

• Speech motor control is complex, involving hierarchical control and sensory feedback.
• Existing models often lack detailed integration of internal state estimation and feedback processing.

Purpose of the Study:

• To introduce the Feedback-Aware Control of Tasks in Speech (FACTS) model.
• To investigate the role of auditory and somatosensory feedback in speech motor control using a computational approach.

Main Methods:

• Developed a hierarchical state feedback control architecture for a simulated vocal tract.
• Integrated forward models for predicting vocal tract state and sensory feedback.
• Simulated noise and feedback alterations to test model robustness and sensory reliance.

Main Results:

• The FACTS model produced intelligible speech and replicated human speech characteristics.
• The model demonstrated robustness to sensory and motor noise.
• It showed differential impacts of auditory versus somatosensory feedback loss and adaptation to altered feedback.

Conclusions:

• The FACTS model provides a framework for understanding speech motor control, highlighting the interplay of sensory feedback.
• It elucidates how sensory acuity may mediate the trade-off between auditory and somatosensory feedback reliance.
• This work has implications for understanding speech disorders and developing advanced speech technology.