Virtuoso: Massive Multilingual Speech-Text Joint Semi-Supervised Learning for Text-to-Speech

• Published in: ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) pp. 1 - 5
• Main Authors: Saeki, Takaaki, Zen, Heiga, Chen, Zhehuai, Morioka, Nobuyuki, Wang, Gary, Zhang, Yu, Bapna, Ankur, Rosenberg, Andrew, Ramabhadran, Bhuvana
• Format: Conference Proceeding
• Language: English
• Published: IEEE 04-06-2023
• Subjects: Acoustics; Data models; massive multilingual pretraining; Multilingual text-to-speech synthesis; Semisupervised learning; Signal processing; Speech synthesis; speech-text semi-supervised joint learning; Task analysis; Training
• DOI: 10.1109/ICASSP49357.2023.10095702
• EISSN: 2379-190X

This paper proposes Virtuoso, a massively multilingual speech-text joint semi-supervised learning framework for text-to-speech synthesis (TTS) models. Existing multilingual TTS typically supports tens of languages, which are a small fraction of the thousands of languages in the world. One difficulty to scale multilingual TTS to hundreds of languages is collecting high-quality speech-text paired data in low-resource languages. This study extends Maestro, a speech-text joint pretraining framework for automatic speech recognition (ASR), to speech generation tasks. To train a TTS model from various types of speech and text data, different training schemes are designed to handle supervised (paired TTS and ASR data) and unsupervised (untranscribed speech and unspoken text) datasets. Experimental evaluation shows that 1) multilingual TTS models trained on Virtuoso can achieve significantly better naturalness and intelligibility than baseline ones in seen languages, and 2) they can synthesize reasonably intelligible and naturally sounding speech for unseen languages where no high-quality paired TTS data is available.