JETS: Jointly Training FastSpeech2 and HiFi-GAN for End to End Text to Speech

TL;DR

This paper introduces an end-to-end TTS model that jointly trains FastSpeech2 and HiFi-GAN with an alignment module, simplifying the training process and improving synthesis quality without needing fine-tuning or external alignment tools.

Contribution

The novel joint training framework integrates FastSpeech2 and HiFi-GAN with an alignment learning objective, eliminating the need for separate training stages and external alignment tools.

Findings

Outperforms state-of-the-art TTS models on LJSpeech in subjective MOS scores.

Simplifies training pipeline by removing fine-tuning and external alignment dependencies.

Achieves higher synthesis quality with end-to-end joint training.

Abstract

In neural text-to-speech (TTS), two-stage system or a cascade of separately learned models have shown synthesis quality close to human speech. For example, FastSpeech2 transforms an input text to a mel-spectrogram and then HiFi-GAN generates a raw waveform from a mel-spectogram where they are called an acoustic feature generator and a neural vocoder respectively. However, their training pipeline is somewhat cumbersome in that it requires a fine-tuning and an accurate speech-text alignment for optimal performance. In this work, we present end-to-end text-to-speech (E2E-TTS) model which has a simplified training pipeline and outperforms a cascade of separately learned models. Specifically, our proposed model is jointly trained FastSpeech2 and HiFi-GAN with an alignment module. Since there is no acoustic feature mismatch between training and inference, it does not requires fine-tuning. Furthermore, we remove dependency on an external speech-text alignment tool by adopting an alignment learning objective in our joint training framework. Experiments on LJSpeech corpus shows that the proposed model outperforms publicly available, state-of-the-art implementations of ESPNet2-TTS on subjective evaluation (MOS) and some objective evaluations.