Generative Adversarial Network-Based Glottal Waveform Model for Statistical Parametric Speech Synthesis

By Bajibabu Bollepalli, Lauri Juvela, Paavo Alku. 2017

50 citations recorded by the source corpus.

Recent studies have shown that text-to-speech synthesis quality can be improved by using glottal vocoding. This refers to vocoders that parameterize speech into two parts, the glottal excitation and vocal tract, that occur in the human speech production apparatus. Current glottal vocoders generate the glottal excitation waveform by using deep neural networks (DNNs). However, the squared error-based training of the present glottal excitation models is limited to generating conditional average waveforms, which fails to capture the stochastic variation of the waveforms. As a result, shaped noise is added as post-processing. In this study, we propose a new method for predicting glottal waveforms by generative adversarial networks (GANs). GANs are generative models that aim to embed the data distribution in a latent space, enabling generation of new instances very similar to the original by randomly sampling the latent distribution. The glottal pulses generated by GANs show a stochastic component similar to natural glottal pulses. In our experiments, we compare synthetic speech generated using glottal waveforms produced by both DNNs and GANs. The results show that the newly proposed GANs achieve synthesis quality comparable to that of widely-used DNNs, without using an additive noise component.

Subfields: Artificial Intelligence, Physiology, 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

In Physiology

In Signal Processing

Citations in the corpus

Parallel Wavegan: A Fast Waveform Generation Model Based on Generative Adversarial Networks with Multi-Resolution Spectrogram (2019, methodology): “3 Our method adopts least-squares GANs thanks to its training stability [19–22].”
Probability density distillation with generative adversarial networks for high-quality parallel waveform generation (2019, methodology): “2 This framework adopts a least-squares GANs thanks to its stability during the training process [22, 24, 27, 28].”
Improving Unsupervised Style Transfer in end-to-end Speech Synthesis with end-to-end Speech Recognition (2018, methodology): “Generative adversarial network (GAN) [23] is used to generate glottal waveform, which can solve the problem that the model can only produce average waveforms when using only squared error loss [24, 25].”
Speaker-independent raw waveform model for glottal excitation (2018, methodology): “More recently, generative adversarial networks have been applied to the task to enable stochastic generation [12, 13], but these models are still constrained by the pitch synchronous windowing scheme.”
Gaussian Lpcnet for Multisample Speech Synthesis (2020, unknown): “[5, 12, 13, 14]).”
GlotNet—A Raw Waveform Model for the Glottal Excitation in Statistical Parametric Speech Synthesis (2019, methodology): “Recently, an alternative non-parametric method of generating stochastic components has been proposed, utilizing generative adversarial networks (GANs) [45], [46].”
Statistical Parametric Speech Synthesis Incorporating Generative Adversarial Networks (2017, methodology): “This is a pioneering method of GAN-based speech synthesis and can be applied not only statistical parametric approaches but also the ones such as glottal waveform synthesis [53].”