TTS-by-TTS: TTS-Driven Data Augmentation for Fast and High-Quality Speech Synthesis

By Min-Jae Hwang, Ryuichi Yamamoto, Eunwoo Song, Jae-Min Kim. 2020

36 citations recorded by the source corpus.

In this paper, we propose a text-to-speech (TTS)-driven data augmentation method for improving the quality of a non-autoregressive (AR) TTS system. Recently proposed non-AR models, such as FastSpeech 2, have successfully achieved fast speech synthesis system. However, their quality is not satisfactory, especially when the amount of training data is insufficient. To address this problem, we propose an effective data augmentation method using a well-designed AR TTS system. In this method, large-scale synthetic corpora including text-waveform pairs with phoneme duration are generated by the AR TTS system, and then used to train the target non-AR model. Perceptual listening test results showed that the proposed method significantly improved the quality of the non-AR TTS system. In particular, we augmented five hours of a training database to 179 hours of a synthetic one. Using these databases, our TTS system consisting of a FastSpeech 2 acoustic model with a Parallel WaveGAN vocoder achieved a mean opinion score of 3.74, which is 40% higher than that achieved by the conventional method.

Subfields: Artificial Intelligence, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Xu Tan
2. Tao Qin
3. Tie-Yan Liu
4. Frank K. Soong
5. Ye Jia
6. Sheng Zhao
7. Michelle Tadmor Ramanovich
8. Tal Remez

In Signal Processing

1. Xu Tan
2. Tao Qin
3. Tie-Yan Liu
4. Frank K. Soong
5. Ye Jia
6. Michelle Tadmor Ramanovich
7. Tal Remez
8. Roi Pomerantz

Citations in the corpus

• A Survey on Neural Speech Synthesis (2021, background): “Research works that scale to unseen domain can alleviate this issue, such as increasing the amount and diversity of the training data [127], adopting relative position encoding to support long sequence unseen in training [17, 423], etc.”
• Translatotron 2: High-quality direct speech-to-speech translation with voice preservation (2021, result): “However, it caused a significant regression on the naturalness of the predicted translation speech, which is consistent with the observation in TTS in Shen et al. (2020); Peng et al. (2020); Hwang et al. (2021), etc., suggesting more exploration is needed on this direction.”