Paper Abstract and Keywords |
Presentation |
2019-02-28 13:55
Model Compression for ECG Signals Outlier Detection Hardware trained by Sparse Robust Deep Autoencoder Naoto Soga, Shimpei Sato, Hiroki Nakahara (Titech) VLD2018-114 HWS2018-77 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
In recent years, portable electrocardiographs and wearable devices have begun to spread so that electrocar- diogram (ECG) signals can be recorded in everyday life. Many types of research have been conducted to use machine learning techniques, including deep learning techniques, to analyze ECG data. However, deep learning models often have too many parameters to implement on mobile hardware. In this research, we propose a method to implement an ECG outlier detector using an autoencoder, which is based on a neural network, in a small built-in device. As a learning method, Robust Deep Autoencoder, one of the unsupervised learning method, was used. A sparseness technique was applied to the autoencoder, and the number of parameters was reduced. Also, Weight Sharing was applied to the obtained weight parameters. With Weight Sharing, the capacity occupied by the weight parameters was reduced by 75% compared with the case where only the sparseness technique was applied. We implemented the obtained Autoencoder on an FPGA. The speed is 20.2 times faster, and 182 times more power efficient than CPUs. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
outlier detection / autoencoder / sparse network / K-Means / unsupervised learning / FPGA / / |
Reference Info. |
IEICE Tech. Rep., vol. 118, no. 457, VLD2018-114, pp. 127-132, Feb. 2019. |
Paper # |
VLD2018-114 |
Date of Issue |
2019-02-20 (VLD, HWS) |
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 |
VLD2018-114 HWS2018-77 |
Conference Information |
Committee |
HWS VLD |
Conference Date |
2019-02-27 - 2019-03-02 |
Place (in Japanese) |
(See Japanese page) |
Place (in English) |
Okinawa Ken Seinen Kaikan |
Topics (in Japanese) |
(See Japanese page) |
Topics (in English) |
Design Technology for System-on-Silicon, Hardware Security, etc. |
Paper Information |
Registration To |
VLD |
Conference Code |
2019-02-HWS-VLD |
Language |
Japanese |
Title (in Japanese) |
(See Japanese page) |
Sub Title (in Japanese) |
(See Japanese page) |
Title (in English) |
Model Compression for ECG Signals Outlier Detection Hardware trained by Sparse Robust Deep Autoencoder |
Sub Title (in English) |
|
Keyword(1) |
outlier detection |
Keyword(2) |
autoencoder |
Keyword(3) |
sparse network |
Keyword(4) |
K-Means |
Keyword(5) |
unsupervised learning |
Keyword(6) |
FPGA |
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 |
2019-02-28 13:55:00 |
Presentation Time |
25 minutes |
Registration for |
VLD |
Paper # |
VLD2018-114, HWS2018-77 |
Volume (vol) |
vol.118 |
Number (no) |
no.457(VLD), no.458(HWS) |
Page |
pp.127-132 |
#Pages |
6 |
Date of Issue |
2019-02-20 (VLD, HWS) |
|