DurIAN-E: Duration Informed Attention Network For Expressive Text-to-Speech Synthesis

TL;DR

This paper presents DurIAN-E, an advanced neural TTS model that enhances expressiveness and quality by integrating duration-informed attention, SwishRNN Transformer encoders, style normalization, and diffusion-based denoising.

Contribution

It introduces DurIAN-E, combining novel encoder architectures, style normalization, and diffusion models for superior expressive TTS synthesis.

Findings

Outperforms state-of-the-art models in MOS scores

Achieves higher preference test results

Enhances expressiveness and naturalness of synthetic speech

Abstract

This paper introduces an improved duration informed attention neural network (DurIAN-E) for expressive and high-fidelity text-to-speech (TTS) synthesis. Inherited from the original DurIAN model, an auto-regressive model structure in which the alignments between the input linguistic information and the output acoustic features are inferred from a duration model is adopted. Meanwhile the proposed DurIAN-E utilizes multiple stacked SwishRNN-based Transformer blocks as linguistic encoders. Style-Adaptive Instance Normalization (SAIN) layers are exploited into frame-level encoders to improve the modeling ability of expressiveness. A denoiser incorporating both denoising diffusion probabilistic model (DDPM) for mel-spectrograms and SAIN modules is conducted to further improve the synthetic speech quality and expressiveness. Experimental results prove that the proposed expressive TTS model in this paper can achieve better performance than the state-of-the-art approaches in both subjective mean opinion score (MOS) and preference tests.