Silent Speech and Emotion Recognition From Vocal Tract Shape Dynamics in Real-time MRI
Artificial Intelligence/Machine Learning
Diversity Equity & Inclusion
DescriptionThis work proposes a novel deep-learning framework that can automatically understand acoustic information present in the vocal tract shaping captured by real-time MRI (rtMRI) during speech production and translate it into text. It also analyzes the variations in the geometry of vocal tract articulators with respect to emotions and genders.