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 2019-09-20 14:00
Quantized Neural Networks Library for Exact Hardware Emulation
Masato Kiyama, Yasuhiro Nakahara, Motoki Amagasaki, Masahiro Iida (Kumamoto Univ.) RECONF2019-33
Abstract (in Japanese) (See Japanese page) 
(in English) Deep neural networks (DNNs) have recently shown outstanding performance in many application domains.
However, it is difficult to run much complex and larger DNN applications on mobile devices due to limited hardware resources.
Quantization is used to reduce the hardware requirements because it uses fewer bits, such as 8-bit fixed points instead of 32-bit floating-point numbers.
DNN librarys use floating-point numbers after quantization because they assume that all data to be calculated in floating-point format.
Therefore, there is mismatch in running hardware emulation exactly.
In this paper, we propose a method of calculation for exact hardware emulation and developed a new DNNs library based on our method.
We show that our library is capable of exact hardware emulation.
Keyword (in Japanese) (See Japanese page) 
(in English) quantization / deep learning / / / / / /  
Reference Info. IEICE Tech. Rep., vol. 119, no. 208, RECONF2019-33, pp. 69-74, Sept. 2019.
Paper # RECONF2019-33 
Date of Issue 2019-09-12 (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 RECONF2019-33

Conference Information
Committee RECONF  
Conference Date 2019-09-19 - 2019-09-20 
Place (in Japanese) (See Japanese page) 
Place (in English) KITAKYUSHU Convention Center 
Topics (in Japanese) (See Japanese page) 
Topics (in English) Reconfigurable Systems, etc. 
Paper Information
Registration To RECONF 
Conference Code 2019-09-RECONF 
Language English (Japanese title is available) 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Quantized Neural Networks Library for Exact Hardware Emulation 
Sub Title (in English)  
Keyword(1) quantization  
Keyword(2) deep learning  
Keyword(3)  
Keyword(4)  
Keyword(5)  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Masato Kiyama  
1st Author's Affiliation Kumamoto University (Kumamoto Univ.)
2nd Author's Name Yasuhiro Nakahara  
2nd Author's Affiliation Kumamoto University (Kumamoto Univ.)
3rd Author's Name Motoki Amagasaki  
3rd Author's Affiliation Kumamoto University (Kumamoto Univ.)
4th Author's Name Masahiro Iida  
4th Author's Affiliation Kumamoto University (Kumamoto 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 ()
Speaker Author-1 
Date Time 2019-09-20 14:00:00 
Presentation Time 20 minutes 
Registration for RECONF 
Paper # RECONF2019-33 
Volume (vol) vol.119 
Number (no) no.208 
Page pp.69-74 
#Pages
Date of Issue 2019-09-12 (RECONF) 


[Return to Top Page]

[Return to IEICE Web Page]


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