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 2019-01-27 11:30
Multimodal Data Augmentation for Visual Speech Recognition using Deep Canonical Correlation Analysis
Masaki Shimonishi, Satoshi Tamura, Satoru Hayamizu (Gifu University) SP2018-60
Abstract (in Japanese) (See Japanese page) 
(in English) This paper proposes ta new data augmentation strategy for deep learning, in which feature vectors in one modality can be used as an additional training data set in the target modality by means of Deep Canonical Correlation Analysis (DCCA). Particularly, this work applies our scheme to visual speech recognition, i.e. lipread- ing, employing audio training feature vectors. A long short-term memory is chosen as a recognition model, which is built using visual training data and augmented data from the audio modality. We evaluated our method using an audio-visual corpus, preparing audio data by adding acoustic noises. Experimental results show applying DCCA enables us to augment training data from the other modalities, and it is turned out that we can further improve the model by implicitly utilize information and knowledge in those modalities.
Keyword (in Japanese) (See Japanese page) 
(in English) lip reading / deep learning / deep canonical correlation analysis / multimodal speech recognition / / / /  
Reference Info. IEICE Tech. Rep., vol. 118, no. 426, SP2018-60, pp. 41-45, Jan. 2019.
Paper # SP2018-60 
Date of Issue 2019-01-19 (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 SP2018-60

Conference Information
Committee SP  
Conference Date 2019-01-26 - 2019-01-27 
Place (in Japanese) (See Japanese page) 
Place (in English) Kanazawa-Harmonie 
Topics (in Japanese) (See Japanese page) 
Topics (in English) Speech Synthesis, Generation, Prosody, Emergency Broadcast, etc. 
Paper Information
Registration To SP 
Conference Code 2019-01-SP 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Multimodal Data Augmentation for Visual Speech Recognition using Deep Canonical Correlation Analysis 
Sub Title (in English)  
Keyword(1) lip reading  
Keyword(2) deep learning  
Keyword(3) deep canonical correlation analysis  
Keyword(4) multimodal speech recognition  
Keyword(5)  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Masaki Shimonishi  
1st Author's Affiliation Gifu University (Gifu University)
2nd Author's Name Satoshi Tamura  
2nd Author's Affiliation Gifu University (Gifu University)
3rd Author's Name Satoru Hayamizu  
3rd Author's Affiliation Gifu University (Gifu University)
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 2019-01-27 11:30:00 
Presentation Time 25 minutes 
Registration for SP 
Paper # SP2018-60 
Volume (vol) vol.118 
Number (no) no.426 
Page pp.41-45 
#Pages
Date of Issue 2019-01-19 (SP) 


[Return to Top Page]

[Return to IEICE Web Page]


The Institute of Electronics, Information and Communication Engineers (IEICE), Japan