PeriodNet: A non-autoregressive waveform generation model with a structure separating periodic and aperiodic components

TL;DR

PeriodNet is a novel non-autoregressive speech waveform generation model that explicitly separates periodic and aperiodic components, improving naturalness and robustness in generated speech, especially for unseen pitch ranges.

Contribution

It introduces a new structure that models periodic and aperiodic components separately without prior decomposition, enhancing speech synthesis quality.

Findings

Improves naturalness of generated speech waveforms.

Effective for pitches outside training data range.

Outperforms existing models in subjective evaluations.

Abstract

We propose PeriodNet, a non-autoregressive (non-AR) waveform generation model with a new model structure for modeling periodic and aperiodic components in speech waveforms. The non-AR waveform generation models can generate speech waveforms parallelly and can be used as a speech vocoder by conditioning an acoustic feature. Since a speech waveform contains periodic and aperiodic components, both components should be appropriately modeled to generate a high-quality speech waveform. However, it is difficult to decompose the components from a natural speech waveform in advance. To address this issue, we propose a parallel model and a series model structure separating periodic and aperiodic components. The features of our proposed models are that explicit periodic and aperiodic signals are taken as input, and external periodic/aperiodic decomposition is not needed in training. Experiments using a singing voice corpus show that our proposed structure improves the naturalness of the generated waveform. We also show that the speech waveforms with a pitch outside of the training data range can be generated with more naturalness.