Towards Parametric Speech Synthesis Using Gaussian-Markov Model of Spectral Envelope and Wavelet-Based Decomposition of F0

TL;DR

This paper introduces a new parametric vocoder combining Gaussian-Markov spectral modeling and wavelet-based F0 decomposition, achieving improved naturalness, robustness, and controllability in speech synthesis compared to traditional methods.

Contribution

It presents a novel vocoder that simplifies waveform generation using Gaussian-Markov spectral modeling and wavelet-based F0 analysis, enhancing controllability and naturalness.

Findings

Outperforms STRAIGHT vocoder in naturalness.

Slightly better than WaveNet in speech quality.

Less effective than WaveRNN but still competitive.

Abstract

Neural network-based Text-to-Speech has significantly improved the quality of synthesized speech. Prominent methods (e.g., Tacotron2, FastSpeech, FastPitch) usually generate Mel-spectrogram from text and then synthesize speech using vocoder (e.g., WaveNet, WaveGlow, HiFiGAN). Compared with traditional parametric approaches (e.g., STRAIGHT and WORLD), neural vocoder based end-to-end models suffer from slow inference speed, and the synthesized speech is usually not robust and lack of controllability. In this work, we propose a novel updated vocoder, which is a simple signal model to train and easy to generate waveforms. We use the Gaussian-Markov model toward robust learning of spectral envelope and wavelet-based statistical signal processing to characterize and decompose F0 features. It can retain the fine spectral envelope and achieve high controllability of natural speech. The experimental results demonstrate that our proposed vocoder achieves better naturalness of reconstructed speech than the conventional STRAIGHT vocoder, slightly better than WaveNet, and somewhat worse than the WaveRNN.