Phoneme-Level BERT for Enhanced Prosody of Text-to-Speech with Grapheme Predictions

TL;DR

This paper introduces a phoneme-level BERT model trained to predict graphemes and phonemes, significantly enhancing the naturalness of synthesized speech in TTS systems, especially on out-of-distribution texts.

Contribution

The paper presents a novel phoneme-level BERT with a grapheme prediction task, improving TTS prosody and naturalness over existing models.

Findings

Significant MOS improvement over SOTA StyleTTS

Enhanced prosody for out-of-distribution texts

Effective phoneme-grapheme joint prediction

Abstract

Large-scale pre-trained language models have been shown to be helpful in improving the naturalness of text-to-speech (TTS) models by enabling them to produce more naturalistic prosodic patterns. However, these models are usually word-level or sup-phoneme-level and jointly trained with phonemes, making them inefficient for the downstream TTS task where only phonemes are needed. In this work, we propose a phoneme-level BERT (PL-BERT) with a pretext task of predicting the corresponding graphemes along with the regular masked phoneme predictions. Subjective evaluations show that our phoneme-level BERT encoder has significantly improved the mean opinion scores (MOS) of rated naturalness of synthesized speech compared with the state-of-the-art (SOTA) StyleTTS baseline on out-of-distribution (OOD) texts.