DiffStyleTTS: Diffusion-based Hierarchical Prosody Modeling for Text-to-Speech with Diverse and Controllable Styles
This addresses the need for more natural and flexible prosodic variations in text-to-speech systems, representing an incremental improvement with specific gains in synthesis speed and control.
The paper tackled the one-to-many mapping problem from text to prosody in text-to-speech by proposing DiffStyleTTS, a diffusion-based model that hierarchically models prosodic features for diverse and controllable styles, resulting in outperforming baselines in naturalness and achieving superior synthesis speed compared to diffusion-based baselines.
Human speech exhibits rich and flexible prosodic variations. To address the one-to-many mapping problem from text to prosody in a reasonable and flexible manner, we propose DiffStyleTTS, a multi-speaker acoustic model based on a conditional diffusion module and an improved classifier-free guidance, which hierarchically models speech prosodic features, and controls different prosodic styles to guide prosody prediction. Experiments show that our method outperforms all baselines in naturalness and achieves superior synthesis speed compared to three diffusion-based baselines. Additionally, by adjusting the guiding scale, DiffStyleTTS effectively controls the guidance intensity of the synthetic prosody.