paper-with-me

Papers

Wave-U-Net Discriminator: Fast and Lightweight Discriminator for Generative Adversarial Network-Based Speech Synthesis

2023-03-24 · Takuhiro Kaneko, Hirokazu Kameoka, Kou Tanaka, Shogo Seki

In speech synthesis, a generative adversarial network (GAN), training a generator (speech synthesizer) and a discriminator in a min-max game, is widely used to improve speech quality. An ensemble of discriminators is commonly used in recent neural vocoders (e.g., HiFi-GAN) and end-to-end text-to-speech (TTS) systems (e.g., VITS) to scrutinize waveforms from multiple perspectives. Such discriminators allow synthesized speech to adequately approach real speech; however, they require an increase in the model size and computation time according to the increase in the number of discriminators. Alternatively, this study proposes a Wave-U-Net discriminator, which is a single but expressive discriminator with Wave-U-Net architecture. This discriminator is unique; it can assess a waveform in a sample-wise manner with the same resolution as the input signal, while extracting multilevel features via an encoder and decoder with skip connections. This architecture provides a generator with sufficiently rich information for the synthesized speech to be closely matched to the real speech. During the experiments, the proposed ideas were applied to a representative neural vocoder (HiFi-GAN) and an end-to-end TTS system (VITS). The results demonstrate that the proposed models can achieve comparable speech quality with a 2.31 times faster and 14.5 times more lightweight discriminator when used in HiFi-GAN and a 1.90 times faster and 9.62 times more lightweight discriminator when used in VITS. Audio samples are available at https://www.kecl.ntt.co.jp/people/kaneko.takuhiro/projects/waveunetd/.

📄 PDF Abstract BibTeX arXiv:2303.13909

Code (0)

등록된 구현이 없습니다.

Tasks

Generative Adversarial NetworkSpeech Synthesistext-to-speechText to Speech

Methods 이 논문이 사용한 방법론

HiFi-GAN HiFi-GAN is a generative adversarial network for speech synthesis. HiFi-GAN consists of one generator and two discriminators: multi-scale and multi-period discriminators. The…

Similar Papers 제목 키워드 기반

Parallel waveform synthesis based on generative adversarial networks with voicing-aware conditional discriminators

2020-10-27 · Ryuichi Yamamoto, Eunwoo Song, Min-Jae Hwang, Jae-Min Kim

This paper proposes voicing-aware conditional discriminators for Parallel WaveGAN-based waveform synthesis systems. In this framework, we adopt a projection-based conditioning method that can significantly improve the di…

text-to-speechText to Speech

Avocodo: Generative Adversarial Network for Artifact-free Vocoder

2022-06-27 · Taejun Bak, Junmo Lee, Hanbin Bae, Jinhyeok Yang 외

Neural vocoders based on the generative adversarial neural network (GAN) have been widely used due to their fast inference speed and lightweight networks while generating high-quality speech waveforms. Since the perceptu…

Generative Adversarial Network

Training Generative Adversarial Network-Based Vocoder with Limited Data Using Augmentation-Conditional Discriminator

2024-03-25 · Takuhiro Kaneko, Hirokazu Kameoka, Kou Tanaka

A generative adversarial network (GAN)-based vocoder trained with an adversarial discriminator is commonly used for speech synthesis because of its fast, lightweight, and high-quality characteristics. However, this data-…

Data AugmentationGenerative Adversarial NetworkSpeech Synthesis

Dynamically Masked Discriminator for Generative Adversarial Networks

2023-06-13 · Wentian Zhang, Haozhe Liu, Bing Li, Jinheng Xie 외

Training Generative Adversarial Networks (GANs) remains a challenging problem. The discriminator trains the generator by learning the distribution of real/generated data. However, the distribution of generated data chang…

Continual Learning

Generative Networks for Precision Enthusiasts

2021-10-22 · Anja Butter, Theo Heimel, Sander Hummerich, Tobias Krebs 외

Generative networks are opening new avenues in fast event generation for the LHC. We show how generative flow networks can reach percent-level precision for kinematic distributions, how they can be trained jointly with a…

Data Augmentation