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 2021-05-21 10:30
A Study on Domain Adaptation for Video Action Classification Utilizing Synthetic Data.
Hana Isoi (Ochanomizu Univ.), Atsuko Takefusa (NII), Hidemoto Nakada (AIST), Masato Oguchi (Ochanomizu Univ.) PRMU2021-5
Abstract (in Japanese) (See Japanese page) 
(in English) The lack of learning data is considered as one of the reasons why the classification accuracies of deep neural networks do not improve. Synthetic data are used for learning to resolve the above issue because they can be generated relatively easier than real data. However, the models trained using synthetic data generally show low classification accuracies for actual data analysis due to domain shift that is differences in data characteristics. To achieve highly accurate video action classification using synthetic video data, we create a realistic synthetic video data set and investigate the video classification accuracies using the data set. The experimental results show that the accuracies can be improved by using data augmentation methods and DNN-based domain adaptation.
Keyword (in Japanese) (See Japanese page) 
(in English) domain adaptation / synthetic data / data augmentation / adversarial learning / video classification / / /  
Reference Info. IEICE Tech. Rep., vol. 121, no. 23, PRMU2021-5, pp. 25-30, May 2021.
Paper # PRMU2021-5 
Date of Issue 2021-05-13 (PRMU) 
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 PRMU2021-5

Conference Information
Committee PRMU IPSJ-CVIM IPSJ-NL  
Conference Date 2021-05-20 - 2021-05-21 
Place (in Japanese) (See Japanese page) 
Place (in English) Online 
Topics (in Japanese) (See Japanese page) 
Topics (in English) Vision & Language 
Paper Information
Registration To PRMU 
Conference Code 2021-05-PRMU-CVIM-NL 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) A Study on Domain Adaptation for Video Action Classification Utilizing Synthetic Data. 
Sub Title (in English)  
Keyword(1) domain adaptation  
Keyword(2) synthetic data  
Keyword(3) data augmentation  
Keyword(4) adversarial learning  
Keyword(5) video classification  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Hana Isoi  
1st Author's Affiliation Ochanomizu University (Ochanomizu Univ.)
2nd Author's Name Atsuko Takefusa  
2nd Author's Affiliation National Institute of Informatics (NII)
3rd Author's Name Hidemoto Nakada  
3rd Author's Affiliation National Institute of Advanced Industrial Science and Technology (AIST)
4th Author's Name Masato Oguchi  
4th Author's Affiliation Ochanomizu University (Ochanomizu Univ.)
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 2021-05-21 10:30:00 
Presentation Time 15 minutes 
Registration for PRMU 
Paper # PRMU2021-5 
Volume (vol) vol.121 
Number (no) no.23 
Page pp.25-30 
#Pages
Date of Issue 2021-05-13 (PRMU) 


[Return to Top Page]

[Return to IEICE Web Page]


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