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 2017-11-07 14:55
DCNN Training with Short Bit Length Format Considering Loss of Trailing Digits
Shin-ichi O'uchi, Hiroshi Fuketa, Ryousei Takano (AIST) CPSY2017-41
Abstract (in Japanese) (See Japanese page) 
(in English) Loss of trailing digits in training deep convolutional neural network (DCNN) is considered to implement training with a format shorter than 16-bit floating format. As a result, we propose an 8-bit training method. The method improves inference accuracy compared to 16-bit floating format in training AlexNet. Simultaneously, the hardware size is reduced by half compared to the 16-bit hardware.
Keyword (in Japanese) (See Japanese page) 
(in English) Deep convolutional neural network (DCNN) / AlexNet / training / floating point arithmetic / accuracy / / /  
Reference Info. IEICE Tech. Rep., vol. 117, no. 278, CPSY2017-41, pp. 7-12, Nov. 2017.
Paper # CPSY2017-41 
Date of Issue 2017-10-31 (CPSY) 
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 CPSY2017-41

Conference Information
Committee VLD DC CPSY RECONF CPM ICD IE IPSJ-SLDM 
Conference Date 2017-11-06 - 2017-11-08 
Place (in Japanese) (See Japanese page) 
Place (in English) Kumamoto-Kenminkouryukan Parea 
Topics (in Japanese) (See Japanese page) 
Topics (in English) Design Gaia 2017 -New Field of VLSI Design- 
Paper Information
Registration To CPSY 
Conference Code 2017-11-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) DCNN Training with Short Bit Length Format Considering Loss of Trailing Digits 
Sub Title (in English)  
Keyword(1) Deep convolutional neural network (DCNN)  
Keyword(2) AlexNet  
Keyword(3) training  
Keyword(4) floating point arithmetic  
Keyword(5) accuracy  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Shin-ichi O'uchi  
1st Author's Affiliation National Institute of AIST (AIST)
2nd Author's Name Hiroshi Fuketa  
2nd Author's Affiliation National Institute of AIST (AIST)
3rd Author's Name Ryousei Takano  
3rd Author's Affiliation National Institute of AIST (AIST)
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 2017-11-07 14:55:00 
Presentation Time 25 minutes 
Registration for CPSY 
Paper # CPSY2017-41 
Volume (vol) vol.117 
Number (no) no.278 
Page pp.7-12 
#Pages
Date of Issue 2017-10-31 (CPSY) 


[Return to Top Page]

[Return to IEICE Web Page]


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