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 2018-08-21 09:15
On Structural Constraint for Convolutional Layers in CNN
Shota Kayamori, Shogo Muramatsu (Niigata Univ.) SIP2018-63
Abstract (in Japanese) (See Japanese page) 
(in English) This work investigates Convolutional Neural Network (CNN) constituted by convolution layers imposed structural constraint. A specific convolutional structure is constructed based on NonSeparable Oversampled Lapped Transforms (NSOLT) and convolutional layers satisfying the linear phase property and Parseval tight frame property is constructed. NSOLT, which the authors proposed in previous works, achieves sparse representation performance equivalent or superior to Sparse K-SVD of Rubinstein et al. Moreover, because of the structural constraint, the number of optimization parameters can be significantly reduced from normal CNN, and convergence acceleration of learning can be expected. In this research, the performance of proposed CNN and usual CNN are compared by using CIFAR-10 data set and the significance of the proposed method is verified.
Keyword (in Japanese) (See Japanese page) 
(in English) CNN / Convolutional layers / Structural constraint / NSOLT / / / /  
Reference Info. IEICE Tech. Rep., vol. 118, no. 182, SIP2018-63, pp. 33-36, Aug. 2018.
Paper # SIP2018-63 
Date of Issue 2018-08-13 (SIP) 
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 SIP2018-63

Conference Information
Committee SIP  
Conference Date 2018-08-20 - 2018-08-21 
Place (in Japanese) (See Japanese page) 
Place (in English) Takushoku Univ. Bunkyo Campus. 
Topics (in Japanese) (See Japanese page) 
Topics (in English)  
Paper Information
Registration To SIP 
Conference Code 2018-08-SIP 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) On Structural Constraint for Convolutional Layers in CNN 
Sub Title (in English)  
Keyword(1) CNN  
Keyword(2) Convolutional layers  
Keyword(3) Structural constraint  
Keyword(4) NSOLT  
Keyword(5)  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Shota Kayamori  
1st Author's Affiliation Niigata University (Niigata Univ.)
2nd Author's Name Shogo Muramatsu  
2nd Author's Affiliation Niigata University (Niigata 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 ()
Speaker Author-1 
Date Time 2018-08-21 09:15:00 
Presentation Time 25 minutes 
Registration for SIP 
Paper # SIP2018-63 
Volume (vol) vol.118 
Number (no) no.182 
Page pp.33-36 
#Pages
Date of Issue 2018-08-13 (SIP) 


[Return to Top Page]

[Return to IEICE Web Page]


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