Paper Abstract and Keywords |
Presentation |
2021-12-22 13:30
[Poster Presentation]
Improved voice quality due to multi-speaker learning with WaveNet vocoder Satoshi Yoshida, Shingo Uenohara, Ken'ichi Furuya (Oita Univ.) EA2021-57 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
In recent years, speech synthesis and voice quality conversion techniques using neural networks have attracted much attention and are capable of synthesizing speech with high naturalness. In order to train a neural vocoder such as WaveNet vocoder, a large amount of speech of the target speaker is required. So far, research has been conducted on training speech of multiple speakers (speakers other than the target speaker). However, there is a problem that the speech quality of the synthesized speech of the WaveNet vocoder trained with the speech of multiple speakers is degraded compared with that trained with the speech of the target speaker. In this study, we propose a method of adding a new convolutional layer to the conventionalWaveNet in order to improve the speech quality of theWaveNet vocoder based on multi-speaker learning. We also confirm whether the speech quality can be improved by fine tuning with a small amount of training data of the target speaker. From the results of evaluation experiments, we confirm that the proposed method improves the speech quality compared to the conventional method. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
Vocoder / WaveNet / Speech Synthesis / Deep Learning / / / / |
Reference Info. |
IEICE Tech. Rep., vol. 121, no. 311, EA2021-57, pp. 1-6, Dec. 2021. |
Paper # |
EA2021-57 |
Date of Issue |
2021-12-15 (EA) |
ISSN |
Online edition: ISSN 2432-6380 |
Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
Download PDF |
EA2021-57 |
Conference Information |
Committee |
EA US |
Conference Date |
2021-12-22 - 2021-12-23 |
Place (in Japanese) |
(See Japanese page) |
Place (in English) |
Sojo University |
Topics (in Japanese) |
(See Japanese page) |
Topics (in English) |
[Joint Meeting on Acoustics and Ultrasonics Subsociety] Engineering/Electro Acoustics, Ultrasonics, etc. |
Paper Information |
Registration To |
EA |
Conference Code |
2021-12-EA-US |
Language |
Japanese |
Title (in Japanese) |
(See Japanese page) |
Sub Title (in Japanese) |
(See Japanese page) |
Title (in English) |
Improved voice quality due to multi-speaker learning with WaveNet vocoder |
Sub Title (in English) |
|
Keyword(1) |
Vocoder |
Keyword(2) |
WaveNet |
Keyword(3) |
Speech Synthesis |
Keyword(4) |
Deep Learning |
Keyword(5) |
|
Keyword(6) |
|
Keyword(7) |
|
Keyword(8) |
|
1st Author's Name |
Satoshi Yoshida |
1st Author's Affiliation |
Oita University (Oita Univ.) |
2nd Author's Name |
Shingo Uenohara |
2nd Author's Affiliation |
Oita University (Oita Univ.) |
3rd Author's Name |
Ken'ichi Furuya |
3rd Author's Affiliation |
Oita University (Oita Univ.) |
4th Author's Name |
|
4th Author's Affiliation |
() |
5th Author's Name |
|
5th Author's Affiliation |
() |
6th Author's Name |
|
6th Author's Affiliation |
() |
7th Author's Name |
|
7th Author's Affiliation |
() |
8th Author's Name |
|
8th Author's Affiliation |
() |
9th Author's Name |
|
9th Author's Affiliation |
() |
10th Author's Name |
|
10th Author's Affiliation |
() |
11th Author's Name |
|
11th Author's Affiliation |
() |
12th Author's Name |
|
12th Author's Affiliation |
() |
13th Author's Name |
|
13th Author's Affiliation |
() |
14th Author's Name |
|
14th Author's Affiliation |
() |
15th Author's Name |
|
15th Author's Affiliation |
() |
16th Author's Name |
|
16th Author's Affiliation |
() |
17th Author's Name |
|
17th Author's Affiliation |
() |
18th Author's Name |
|
18th Author's Affiliation |
() |
19th Author's Name |
|
19th Author's Affiliation |
() |
20th Author's Name |
|
20th Author's Affiliation |
() |
Speaker |
Author-1 |
Date Time |
2021-12-22 13:30:00 |
Presentation Time |
120 minutes |
Registration for |
EA |
Paper # |
EA2021-57 |
Volume (vol) |
vol.121 |
Number (no) |
no.311 |
Page |
pp.1-6 |
#Pages |
6 |
Date of Issue |
2021-12-15 (EA) |
|