Semi-Supervised Training for Improving Data Efficiency in End-to-End Speech Synthesis

TL;DR

This paper introduces a semi-supervised training method for end-to-end speech synthesis that leverages large unpaired text and speech data, significantly reducing the need for expensive paired datasets.

Contribution

It proposes a novel semi-supervised framework that uses unpaired data for pre-training and fine-tuning, improving data efficiency in Tacotron-based speech synthesis.

Findings

Achieves intelligible speech synthesis with less than half an hour of paired data

Utilizes unpaired text data through word embedding conditioning

Pre-trains decoder with unpaired speech corpus

Abstract

Although end-to-end text-to-speech (TTS) models such as Tacotron have shown excellent results, they typically require a sizable set of high-quality <text, audio> pairs for training, which are expensive to collect. In this paper, we propose a semi-supervised training framework to improve the data efficiency of Tacotron. The idea is to allow Tacotron to utilize textual and acoustic knowledge contained in large, publicly-available text and speech corpora. Importantly, these external data are unpaired and potentially noisy. Specifically, first we embed each word in the input text into word vectors and condition the Tacotron encoder on them. We then use an unpaired speech corpus to pre-train the Tacotron decoder in the acoustic domain. Finally, we fine-tune the model using available paired data. We demonstrate that the proposed framework enables Tacotron to generate intelligible speech using less than half an hour of paired training data.