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 2024-10-17 14:00
Consideration of incremental learning in the latent variable space
Chika Obata, Kenya Jin'no (Tokyo City Univ.) CAS2024-37 NLP2024-67
Abstract (in Japanese) (See Japanese page) 
(in English) In this article, we examine the methods and problems of incremental learning in classifying deep learning models.
In conventional classifying tasks, the model is trained using all the class data at once, but in order to mimic the flexibility of the human learning process, we focus on the possibility of ‘additional learning’, where when a new class is added, the existing model is trained on the new class.
A method of this kind of incremental learning is the Direct ONE-shot learning (DONE) method, which has been proposed recently. In this study, in order to examine the effectiveness of the DONE method, we added new class data to an existing deep neural network model and conducted an experiment to check how the interference in class identification after the addition would be. Using the CIFAR-10 dataset, we conduct an experiment in which only a specific class is added as a new class for training, and examined how the distribution of feature values changed. Furthermore, in order to determine whether the feature values of images are shape or colour information, we examine the accuracy of classification when images with specific channel information extracted from input colour images are input.
Keyword (in Japanese) (See Japanese page) 
(in English) deep learning / incremental learning / catastrophic forgetting / class identification / feature distribution / / /  
Reference Info. IEICE Tech. Rep., vol. 124, no. 208, NLP2024-67, pp. 52-57, Oct. 2024.
Paper # NLP2024-67 
Date of Issue 2024-10-10 (CAS, NLP) 
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 CAS2024-37 NLP2024-67

Conference Information
Committee NLP CAS  
Conference Date 2024-10-17 - 2024-10-18 
Place (in Japanese) (See Japanese page) 
Place (in English) Information Center, Tottori University 
Topics (in Japanese) (See Japanese page) 
Topics (in English) Nonlinear Problems, Circuits and Systems, etc. 
Paper Information
Registration To NLP 
Conference Code 2024-10-NLP-CAS 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Consideration of incremental learning in the latent variable space 
Sub Title (in English)  
Keyword(1) deep learning  
Keyword(2) incremental learning  
Keyword(3) catastrophic forgetting  
Keyword(4) class identification  
Keyword(5) feature distribution  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Chika Obata  
1st Author's Affiliation Tokyo City University (Tokyo City Univ.)
2nd Author's Name Kenya Jin'no  
2nd Author's Affiliation Tokyo City University (Tokyo City Univ.)
3rd Author's Name  
3rd Author's Affiliation ()
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 2024-10-17 14:00:00 
Presentation Time 20 minutes 
Registration for NLP 
Paper # CAS2024-37, NLP2024-67 
Volume (vol) vol.124 
Number (no) no.207(CAS), no.208(NLP) 
Page pp.52-57 
#Pages 6 
Date of Issue 2024-10-10 (CAS, NLP) 

[Return to Top Page]

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