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 2024-03-13 16:40
A Plug-and-Play Module for Enhancing Fault-Tolerant Distributed Inference Based on Gaussian Dropout
Hou Zhangcheng, Ohtsuki Tomoaki (KU) RCS2023-267
Abstract (in Japanese) (See Japanese page) 
(in English) Distributed inference (DI) in the Internet of Things (IoT) is becoming increasingly important as the demand for AI applications grows. When unreliable links are used for IoT transmission, missing data can adversely affect inference accuracy. Therefore, we designed a plug-and-play module to enhance the robustness of DI systems. The dropout we utilize is no longer a fixed value but is sampled from a Gaussian distribution, making it better adapted to lossy networks. Additionally, the convolutional layer of the module can learn and preserve fault-tolerant inference. This method uses freeze training and does not require retraining of the original deep neural network, thus significantly reducing the number of parameters to be trained and avoiding catastrophic forgetting. We incorporate various designs for the Gaussian dropout layer to further enhance its effectiveness at low packet loss rates. The module only needs to be attached to the cut-points of the original deep neural network, making it very easy and fast to deploy. Experimental results demonstrate that our plug-and-play module can be adapted to different deep neural networks and various lossy network scenarios, achieving an average accuracy improvement in all tests. Results show that our plug-and-play module significantly enhances the fault-tolerant inference of the system.
Keyword (in Japanese) (See Japanese page) 
(in English) Internet of Things / error-tolerant / distributed inference / deep learning / / / /  
Reference Info. IEICE Tech. Rep., vol. 123, no. 434, RCS2023-267, pp. 77-82, March 2024.
Paper # RCS2023-267 
Date of Issue 2024-03-06 (RCS) 
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 RCS2023-267

Conference Information
Committee RCS SR SRW  
Conference Date 2024-03-13 - 2024-03-15 
Place (in Japanese) (See Japanese page) 
Place (in English) The University of Tokyo (Hongo Campus), and online 
Topics (in Japanese) (See Japanese page) 
Topics (in English) Mobile Communication Workshop 
Paper Information
Registration To RCS 
Conference Code 2024-03-RCS-SR-SRW 
Language English 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) A Plug-and-Play Module for Enhancing Fault-Tolerant Distributed Inference Based on Gaussian Dropout 
Sub Title (in English)  
Keyword(1) Internet of Things  
Keyword(2) error-tolerant  
Keyword(3) distributed inference  
Keyword(4) deep learning  
Keyword(5)  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Hou Zhangcheng  
1st Author's Affiliation Keio University (KU)
2nd Author's Name Ohtsuki Tomoaki  
2nd Author's Affiliation Keio University (KU)
3rd Author's Name  
3rd Author's Affiliation ()
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 ()
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 2024-03-13 16:40:00 
Presentation Time 25 minutes 
Registration for RCS 
Paper # RCS2023-267 
Volume (vol) vol.123 
Number (no) no.434 
Page pp.77-82 
#Pages 6 
Date of Issue 2024-03-06 (RCS) 

[Return to Top Page]

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