Information: Join today and make your research activities more affordable! Technical workshop participation fees and annual registration fees are available at member rates.
Notice: [Important] Announcement of Changes to Registration Fee Payment and Manuscript Upload Procedures for IEICE Technical Meetings
IEICE Technical Committee Submission System
Conference Paper's Information		 Online Proceedings
[Sign in]
Tech. Rep. Archives
 Go Top Page Go Previous   [Japanese] / [English] 

Paper Abstract and Keywords
Presentation 2018-03-19 10:25
Non-parallel and Many-to-Many Voice Conversion Using Variational Autoencoder Conditioned by Phonetic Posteriorgrams and d-vectors
Yuki Saito (NTT/Univ. of Tokyo), Yusuke Ijima, Kyosuke Nishida (NTT), Shinnosuke Takamichi (Univ. of Tokyo) EA2017-105 SIP2017-114 SP2017-88
Abstract (in Japanese) (See Japanese page) 
(in English) This paper proposes novel frameworks for non-parallel and many-to-many voice conversion (VC) using variational autoencoders (VAEs). In conventional VAE-based VC, converted speech quality is significantly degraded due to an over-regularization of latent variables representing phonetic contents. To overcome the issue, this paper proposes a VAE-based non-parallel VC conditioned by not only the speaker codes but also phonetic posteriorgrams (PPGs) predicted from pre-trained speech recognition models. This paper also extends the conventional VC to many-to-many VC that can convert arbitrary speakers’ characteristics into another ones. We compare two methods to realize this: 1) speaker code adaptation, and 2) the use of $d$-vectors obtained by using pre-trained speaker verification models. Experimental results demonstrate that 1) PPGs successfully improve converted speech quality, and 2) both speaker codes and $d$-vectors can be adopted to the VAE-based non-parallel and many-to-many VC.
Keyword (in Japanese) (See Japanese page) 
(in English) non-parallel voice conversion / many-to-many voice conversion / variational autoencoders / phonetic posteriorgrams / d-vectors / / /  
Reference Info. IEICE Tech. Rep., vol. 117, no. 517, SP2017-88, pp. 21-26, March 2018.
Paper # SP2017-88 
Date of Issue 2018-03-12 (EA, SIP, SP) 
ISSN Print edition: ISSN 0913-5685    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 EA2017-105 SIP2017-114 SP2017-88

Conference Information
Committee SIP EA SP MI  
Conference Date 2018-03-19 - 2018-03-20 
Place (in Japanese) (See Japanese page) 
Place (in English)  
Topics (in Japanese) (See Japanese page) 
Topics (in English) Speech, Engineering/Electro Acoustics, Signal Processing, and Related Topics [SIP, EA, SP]/ Medical Image Engineering, Analysis, Recognition, etc. [MI] 
Paper Information
Registration To SP 
Conference Code 2018-03-SIP-EA-SP-MI 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Non-parallel and Many-to-Many Voice Conversion Using Variational Autoencoder Conditioned by Phonetic Posteriorgrams and d-vectors 
Sub Title (in English)  
Keyword(1) non-parallel voice conversion  
Keyword(2) many-to-many voice conversion  
Keyword(3) variational autoencoders  
Keyword(4) phonetic posteriorgrams  
Keyword(5) d-vectors  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Yuki Saito  
1st Author's Affiliation NTT Media Intelligence Laboratories, NTT Corporation/The University of Tokyo (NTT/Univ. of Tokyo)
2nd Author's Name Yusuke Ijima  
2nd Author's Affiliation NTT Media Intelligence Laboratories, NTT Corporation (NTT)
3rd Author's Name Kyosuke Nishida  
3rd Author's Affiliation NTT Media Intelligence Laboratories, NTT Corporation (NTT)
4th Author's Name Shinnosuke Takamichi  
4th Author's Affiliation The University of Tokyo (Univ. of Tokyo)
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 ()
21st Author's Name  
21st Author's Affiliation ()
22nd Author's Name  
22nd Author's Affiliation ()
23rd Author's Name  
23rd Author's Affiliation ()
24th Author's Name  
24th Author's Affiliation ()
25th Author's Name  
25th Author's Affiliation ()
26th Author's Name / /
26th Author's Affiliation ()
()
27th Author's Name / /
27th Author's Affiliation ()
()
28th Author's Name / /
28th Author's Affiliation ()
()
29th Author's Name / /
29th Author's Affiliation ()
()
30th Author's Name / /
30th Author's Affiliation ()
()
31st Author's Name / /
31st Author's Affiliation ()
()
32nd Author's Name / /
32nd Author's Affiliation ()
()
33rd Author's Name / /
33rd Author's Affiliation ()
()
34th Author's Name / /
34th Author's Affiliation ()
()
35th Author's Name / /
35th Author's Affiliation ()
()
36th Author's Name / /
36th Author's Affiliation ()
()
Speaker Author-1 
Date Time 2018-03-19 10:25:00 
Presentation Time 25 minutes 
Registration for SP 
Paper # EA2017-105, SIP2017-114, SP2017-88 
Volume (vol) vol.117 
Number (no) no.515(EA), no.516(SIP), no.517(SP) 
Page pp.21-26 
#Pages 6 
Date of Issue 2018-03-12 (EA, SIP, SP) 

[Return to Top Page]

[Return to IEICE Web Page]
The Institute of Electronics, Information and Communication Engineers (IEICE), Japan