Vikram Ramanarayanan

The paper develops a computational model to test cerebellar functions in speech feedback control, finding that internal modeling, timing dynamics, and multimodal integration explain speech abnormalities in cerebellar degeneration.

This paper uses Bayesian inference to estimate parameters of a state feedback control model for vocal pitch in cerebellar ataxia and healthy controls, finding differences in feedback sensitivity and motor response.

This study uses a multimodal dialogue agent to analyze speech and facial metrics in children with ASD, finding correlations between vocal and kinematic metrics and clinical assessments of affect and motor function.

The FACTS model replicates human speech motor control by integrating state estimation and task-based control, showing robustness to noise and feedback loss.