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 2022-03-02 11:35
Study of Method for Improving Speech Intelligibility in Glossectomy Patients by Knowledge Distillation via Lip Features
Kazushi Takashima, Masanobu Abe, Sunao Hara (Okayama Univ.) EA2021-81 SIP2021-108 SP2021-66
Abstract (in Japanese) (See Japanese page) 
(in English) In this paper, we propose a voice conversion method for improving speech intelligibility uttered by glossectomy patients. Because the glossectomy patients remove more than half of their tongue, their speech is less intelligibility compared to healthy persons. In a previous study, a voice conversion method using phoneme labels as auxiliary information in addition to audio information was proposed and it has been shown that the speech intelligibility is greatly improved. However, in actual situations, it is difficult to prepare phoneme labels corresponding to the speech content at the time of conversion. To solve the problem, by introducing knowledge distillation approach, we proposed to train a student model that synthesizes speech without labels, where a model trained using phoneme labels is used as teacher model. Although the performance of the method with knowledge distillation is better than that of the method without knowledge distillation, intelligibility was not imporved enough. Hence, we investigate a method that uses lip information as an additional feature in order to improve the performance of the student model. we extract bottleneck features from lip information and use them as additional input features by using a convolutional autoencoder. In the evaluation experiments, we evaluated the performance of the convolutional autoencoder and the conversion accuracy of each voice conversion method.
Keyword (in Japanese) (See Japanese page) 
(in English) Voice Conversion / Glossectomy Patients / Deep Neural Network / CNN-Autoencoder / Knowledge Distillation / / /  
Reference Info. IEICE Tech. Rep., vol. 121, no. 385, SP2021-66, pp. 108-113, March 2022.
Paper # SP2021-66 
Date of Issue 2022-02-22 (EA, SIP, SP) 
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 EA2021-81 SIP2021-108 SP2021-66

Conference Information
Committee EA SIP SP IPSJ-SLP  
Conference Date 2022-03-01 - 2022-03-02 
Place (in Japanese) (See Japanese page) 
Place (in English)  
Topics (in Japanese) (See Japanese page) 
Topics (in English)  
Paper Information
Registration To SP 
Conference Code 2022-03-EA-SIP-SP-SLP 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Study of Method for Improving Speech Intelligibility in Glossectomy Patients by Knowledge Distillation via Lip Features 
Sub Title (in English)  
Keyword(1) Voice Conversion  
Keyword(2) Glossectomy Patients  
Keyword(3) Deep Neural Network  
Keyword(4) CNN-Autoencoder  
Keyword(5) Knowledge Distillation  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Kazushi Takashima  
1st Author's Affiliation Okayama University (Okayama Univ.)
2nd Author's Name Masanobu Abe  
2nd Author's Affiliation Okayama University (Okayama Univ.)
3rd Author's Name Sunao Hara  
3rd Author's Affiliation Okayama University (Okayama 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 ()
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 2022-03-02 11:35:00 
Presentation Time 25 minutes 
Registration for SP 
Paper # EA2021-81, SIP2021-108, SP2021-66 
Volume (vol) vol.121 
Number (no) no.383(EA), no.384(SIP), no.385(SP) 
Page pp.108-113 
#Pages 6 
Date of Issue 2022-02-22 (EA, SIP, SP) 

[Return to Top Page]

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