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 2020-03-17 09:45
Deep neural network representation and learning of low-rank and sparse approximation -- With application to celiac angiography under free breathing --
Ryohei Miyoshi, Tomoya Sakai (Nagasaki Univ.), Takashi Ohnishi, Hideaki Haneishi (Chiba Univ.) PRMU2019-91
Abstract (in Japanese) (See Japanese page) 
(in English) Low-rank and sparse (L+S) approximation, a.k.a. stable and robust principal component analysis, is known to be suitable for analyzing sequential data simultaneously containing linearly dependent and sparse features.
There remain, however, some challenges in practice.
Hyperparameters ${bf Lambda}_{mbox{scriptsize S}}$ to control the balance between the low-rankness and sparseness must be tuned on real data.
Introducing additional prior knowledge to the L+S structures makes the optimization algorithm computationally intensive.
This work takes a novel approach to improve the L+S approximation by representing its algorithm as a deep neural network (DNN) with trainable hyperparameters ${bf Lambda}_{mbox{scriptsize S}}$.
In order for the DNN to learn via the backpropagation, the singular-value thresholding and soft thresholding in the L+S algorithm, as well as the nuclear and $ell_1$ norms imposing the L+S nature on the DNN outputs, can be implemented as auto-differentiable modules using a deep learning framework.
Introducing the total variation of the sparse components into the loss function for the DNN training, the hyperparameters successfully acquire spatial distribution of the sparse components, which is experimentlly shown to improve artery detction in the application to celiac angiography under free-breathing condition.
Keyword (in Japanese) (See Japanese page) 
(in English) Robust PCA / Hyperparameter optimization / Nuclear loss function / ADMM / / / /  
Reference Info. IEICE Tech. Rep., vol. 119, no. 481, PRMU2019-91, pp. 133-138, March 2020.
Paper # PRMU2019-91 
Date of Issue 2020-03-09 (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 PRMU2019-91

Conference Information
Committee PRMU IPSJ-CVIM  
Conference Date 2020-03-16 - 2020-03-17 
Place (in Japanese) (See Japanese page) 
Place (in English)  
Topics (in Japanese) (See Japanese page) 
Topics (in English)  
Paper Information
Registration To PRMU 
Conference Code 2020-03-PRMU-CVIM 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Deep neural network representation and learning of low-rank and sparse approximation 
Sub Title (in English) With application to celiac angiography under free breathing 
Keyword(1) Robust PCA  
Keyword(2) Hyperparameter optimization  
Keyword(3) Nuclear loss function  
Keyword(4) ADMM  
Keyword(5)  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Ryohei Miyoshi  
1st Author's Affiliation Nagasaki University (Nagasaki Univ.)
2nd Author's Name Tomoya Sakai  
2nd Author's Affiliation Nagasaki University (Nagasaki Univ.)
3rd Author's Name Takashi Ohnishi  
3rd Author's Affiliation Chiba University (Chiba Univ.)
4th Author's Name Hideaki Haneishi  
4th Author's Affiliation Chiba University (Chiba 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 ()
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 2020-03-17 09:45:00 
Presentation Time 15 minutes 
Registration for PRMU 
Paper # PRMU2019-91 
Volume (vol) vol.119 
Number (no) no.481 
Page pp.133-138 
#Pages 6 
Date of Issue 2020-03-09 (PRMU) 

[Return to Top Page]

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