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 2018-12-06 11:20
Hardware implementation of ECG signals outlier detector trained by Sparse Robust Deep Autoencoder
Naoto Soga, Shimpei Sato, Hiroki Nakahara (Titech) RECONF2018-42
Abstract (in Japanese) (See Japanese page) 
(in English) Current ECG outlier detection is rule-based, there are many false positives, and it is necessary to study a new outlier detection method. As a method of detecting an outlier, autoencoder which is one of the models of a neural network is known. In this research, we trained autoencoder using Robust Deep Autoencoder (RDA) which can train an autoencoder without training data. We made the weighting parameters sparse and we reduced the weighting parameters. As a result, about 95% of the weighting parameters were reduced while keeping the F1 score down to about 0.84%. In this research, we also implemented the autoencoder trained by RDA on the FPGA. By adjusting the bit width and the way of storing the weighting parameters, it was possible to reduce the capacity of the weighting parameters by about 95.6% compared to before applying a sparse network to autoencoder.
Keyword (in Japanese) (See Japanese page) 
(in English) outlier detection / autoencoder / a sparse network / FPGA / / / /  
Reference Info. IEICE Tech. Rep., vol. 118, no. 340, RECONF2018-42, pp. 45-50, Dec. 2018.
Paper # RECONF2018-42 
Date of Issue 2018-11-28 (RECONF) 
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 RECONF2018-42

Conference Information
Committee VLD DC CPSY RECONF CPM ICD IE IPSJ-SLDM 
Conference Date 2018-12-05 - 2018-12-07 
Place (in Japanese) (See Japanese page) 
Place (in English) Satellite Campus Hiroshima 
Topics (in Japanese) (See Japanese page) 
Topics (in English) Design Gaia 2018 -New Field of VLSI Design- 
Paper Information
Registration To RECONF 
Conference Code 2018-12-VLD-DC-CPSY-RECONF-CPM-ICD-IE-SLDM-EMB-ARC 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Hardware implementation of ECG signals outlier detector trained by Sparse Robust Deep Autoencoder 
Sub Title (in English)  
Keyword(1) outlier detection  
Keyword(2) autoencoder  
Keyword(3) a sparse network  
Keyword(4) FPGA  
Keyword(5)  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Naoto Soga  
1st Author's Affiliation Tokyo Institute of Technology (Titech)
2nd Author's Name Shimpei Sato  
2nd Author's Affiliation Tokyo Institute of Technology (Titech)
3rd Author's Name Hiroki Nakahara  
3rd Author's Affiliation Tokyo Institute of Technology (Titech)
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 2018-12-06 11:20:00 
Presentation Time 25 minutes 
Registration for RECONF 
Paper # RECONF2018-42 
Volume (vol) vol.118 
Number (no) no.340 
Page pp.45-50 
#Pages
Date of Issue 2018-11-28 (RECONF) 


[Return to Top Page]

[Return to IEICE Web Page]


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