Using VAEs and Normalizing Flows for One-shot Text-To-Speech Synthesis of Expressive Speech
This addresses the challenge of creating personalized and emotional speech synthesis for applications like virtual assistants, though it is incremental as it builds on existing sequence-to-sequence systems.
The paper tackles the problem of generating expressive speech from text using only a one-second reference utterance, achieving a 22% KL-divergence reduction and a 9% relative naturalness improvement over standard neural TTS.
We propose a Text-to-Speech method to create an unseen expressive style using one utterance of expressive speech of around one second. Specifically, we enhance the disentanglement capabilities of a state-of-the-art sequence-to-sequence based system with a Variational AutoEncoder (VAE) and a Householder Flow. The proposed system provides a 22% KL-divergence reduction while jointly improving perceptual metrics over state-of-the-art. At synthesis time we use one example of expressive style as a reference input to the encoder for generating any text in the desired style. Perceptual MUSHRA evaluations show that we can create a voice with a 9% relative naturalness improvement over standard Neural Text-to-Speech, while also improving the perceived emotional intensity (59 compared to the 55 of neutral speech).