StyleTTS 2: Towards Human-Level Text-to-Speech through Style Diffusion and Adversarial Training with Large Speech Language Models
arXiv:2306.07691
Abstract
In this paper, we present StyleTTS 2, a text-to-speech (TTS) model that leverages style diffusion and adversarial training with large speech language models (SLMs) to achieve human-level TTS synthesis. StyleTTS 2 differs from its predecessor by modeling styles as a latent random variable through diffusion models to generate the most suitable style for the text without requiring reference speech, achieving efficient latent diffusion while benefiting from the diverse speech synthesis offered by diffusion models. Furthermore, we employ large pre-trained SLMs, such as WavLM, as discriminators with our novel differentiable duration modeling for end-to-end training, resulting in improved speech naturalness. StyleTTS 2 surpasses human recordings on the single-speaker LJSpeech dataset and matches it on the multispeaker VCTK dataset as judged by native English speakers. Moreover, when trained on the LibriTTS dataset, our model outperforms previous publicly available models for zero-shot speaker adaptation. This work achieves the first human-level TTS on both single and multispeaker datasets, showcasing the potential of style diffusion and adversarial training with large SLMs. The audio demos and source code are available at https://styletts2.github.io/.
arXiv commentNeurIPS 2023
References in corpus (23)
- WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing
- Score-Based Generative Modeling through Stochastic Differential Equations
- HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis
- Elucidating the Design Space of Diffusion-Based Generative Models
- DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps
- A Survey on Neural Speech Synthesis
- Neural Codec Language Models are Zero-Shot Text to Speech Synthesizers
- Non-Attentive Tacotron: Robust and Controllable Neural TTS Synthesis Including Unsupervised Duration Modeling
- Neural Networks Fail to Learn Periodic Functions and How to Fix It
- BigVGAN: A Universal Neural Vocoder with Large-Scale Training
- VQTTS: High-Fidelity Text-to-Speech Synthesis with Self-Supervised VQ Acoustic Feature
- A Survey on Audio Diffusion Models: Text To Speech Synthesis and Enhancement in Generative AI
- A Variational Perspective on Diffusion-Based Generative Models and Score Matching
- NaturalSpeech 2: Latent Diffusion Models are Natural and Zero-Shot Speech and Singing Synthesizers
- NaturalSpeech: End-to-End Text to Speech Synthesis with Human-Level Quality
- FastDiff: A Fast Conditional Diffusion Model for High-Quality Speech Synthesis
- PriorGrad: Improving Conditional Denoising Diffusion Models with Data-Dependent Adaptive Prior
- WavThruVec: Latent speech representation as intermediate features for neural speech synthesis
- ProDiff: Progressive Fast Diffusion Model For High-Quality Text-to-Speech
- DiffGAN-TTS: High-Fidelity and Efficient Text-to-Speech with Denoising Diffusion GANs
- StyleTTS: A Style-Based Generative Model for Natural and Diverse Text-to-Speech Synthesis
- M2-CTTS: End-to-End Multi-scale Multi-modal Conversational Text-to-Speech Synthesis
- FoundationTTS: Text-to-Speech for ASR Customization with Generative Language Model