Saved in:
Bibliographic Details
Main Authors: Valin, Jean-Marc, Mustafa, Ahmed, Büthe, Jan
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2405.21069
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909278898487296
author Valin, Jean-Marc
Mustafa, Ahmed
Büthe, Jan
author_facet Valin, Jean-Marc
Mustafa, Ahmed
Büthe, Jan
contents Neural vocoders are now being used in a wide range of speech processing applications. In many of those applications, the vocoder can be the most complex component, so finding lower complexity algorithms can lead to significant practical benefits. In this work, we propose FARGAN, an autoregressive vocoder that takes advantage of long-term pitch prediction to synthesize high-quality speech in small subframes, without the need for teacher-forcing. Experimental results show that the proposed 600~MFLOPS FARGAN vocoder can achieve both higher quality and lower complexity than existing low-complexity vocoders. The quality even matches that of existing higher-complexity vocoders.
format Preprint
id arxiv_https___arxiv_org_abs_2405_21069
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Very Low Complexity Speech Synthesis Using Framewise Autoregressive GAN (FARGAN) with Pitch Prediction
Valin, Jean-Marc
Mustafa, Ahmed
Büthe, Jan
Audio and Speech Processing
Neural vocoders are now being used in a wide range of speech processing applications. In many of those applications, the vocoder can be the most complex component, so finding lower complexity algorithms can lead to significant practical benefits. In this work, we propose FARGAN, an autoregressive vocoder that takes advantage of long-term pitch prediction to synthesize high-quality speech in small subframes, without the need for teacher-forcing. Experimental results show that the proposed 600~MFLOPS FARGAN vocoder can achieve both higher quality and lower complexity than existing low-complexity vocoders. The quality even matches that of existing higher-complexity vocoders.
title Very Low Complexity Speech Synthesis Using Framewise Autoregressive GAN (FARGAN) with Pitch Prediction
topic Audio and Speech Processing
url https://arxiv.org/abs/2405.21069