This paper presents a novel data augmentation technique for text-to-speech (TTS), that allows to generate new (text, audio) training examples without requiring any additional data. Our goal is to increase diversity of text conditionings available during training. This helps to reduce overfitting, especially in low-resource settings. Our method relies on substituting text and audio fragments in a way that preserves syntactical correctness. We take additional measures to ensure that synthesized speech does not contain artifacts caused by combining inconsistent audio samples. The perceptual evaluations show that our method improves speech quality over a number of datasets, speakers, and TTS architectures. We also demonstrate that it greatly improves robustness of attention-based TTS models.
@article{arxiv.2202.06409,
title = {Distribution augmentation for low-resource expressive text-to-speech},
author = {Mateusz Lajszczak and Animesh Prasad and Arent van Korlaar and Bajibabu Bollepalli and Antonio Bonafonte and Arnaud Joly and Marco Nicolis and Alexis Moinet and Thomas Drugman and Trevor Wood and Elena Sokolova},
journal= {arXiv preprint arXiv:2202.06409},
year = {2022}
}