A Variational Prosody Model for the decomposition and synthesis of speech prosody

The quest for comprehensive generative models of intonation that link linguistic and paralinguistic functions to prosodic forms has been a longstanding challenge of speech communication research. More traditional intonation models have given way to the overwhelming performance of artificial intelligence (AI) techniques for training model-free, end-to-end mappings using millions of tunable parameters. The shift towards machine learning models has nonetheless posed the reverse problem - a compelling need to discover knowledge, to explain, visualise and interpret. Our work bridges between a comprehensive generative model of intonation and state-of-the-art AI techniques. We build upon the modelling paradigm of the Superposition of Functional Contours model and propose a Variational Prosody Model (VPM) that uses a network of deep variational contour generators to capture the context-sensitive variation of the constituent elementary prosodic cliches. We show that the VPM can give insight into the intrinsic variability of these prosodic prototypes through learning a meaningful prosodic latent space representation structure. We also show that the VPM brings improved modelling performance especially when such variability is prominent. In a speech synthesis scenario we believe the model can be used to generate a dynamic and natural prosody contour largely devoid of averaging effects.