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 2020-12-18 14:55
Regularization Using Knowledge Distillation in Learning Small Datasets
Ryota Higashi, Toshikazu Wada (Wakayama Univ.) PRMU2020-61
Abstract (in Japanese) (See Japanese page) 
(in English) Knowledge distillation is a method mainly used for compressing deep learning models, but it has recently gained attention for its effectiveness in learning from small amounts of data as well. In this report, taking the image classification problem as an example, we focused on the fact that the decrease in classification accuracy can be suppressed by distillation when the training data is reduced, and the accuracy varies with a distillation parameter called “temperature”. First, we prepare a teacher model trained on all the training data and then distill it into a student model. In this case, we found that the accuracy of the student model is improved by increasing the temperature, especially when the number of training data is small, and that this effect is not related to the calibration of the teacher model.
Keyword (in Japanese) (See Japanese page) 
(in English) Deep Learning / Knowledge Distillation / Image Classification / Few-Shot Learning / Calibration / / /  
Reference Info. IEICE Tech. Rep., vol. 120, no. 300, PRMU2020-61, pp. 133-138, Dec. 2020.
Paper # PRMU2020-61 
Date of Issue 2020-12-10 (PRMU) 
ISSN Online edition: ISSN 2432-6380
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 PRMU2020-61

Conference Information
Committee PRMU  
Conference Date 2020-12-17 - 2020-12-18 
Place (in Japanese) (See Japanese page) 
Place (in English) Online 
Topics (in Japanese) (See Japanese page) 
Topics (in English) Transfer learning and few shot learning 
Paper Information
Registration To PRMU 
Conference Code 2020-12-PRMU 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Regularization Using Knowledge Distillation in Learning Small Datasets 
Sub Title (in English)  
Keyword(1) Deep Learning  
Keyword(2) Knowledge Distillation  
Keyword(3) Image Classification  
Keyword(4) Few-Shot Learning  
Keyword(5) Calibration  
1st Author's Name Ryota Higashi  
1st Author's Affiliation Wakayama University (Wakayama Univ.)
2nd Author's Name Toshikazu Wada  
2nd Author's Affiliation Wakayama University (Wakayama Univ.)
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 ()
Speaker Author-1 
Date Time 2020-12-18 14:55:00 
Presentation Time 15 minutes 
Registration for PRMU 
Paper # PRMU2020-61 
Volume (vol) vol.120 
Number (no) no.300 
Page pp.133-138 
Date of Issue 2020-12-10 (PRMU) 

[Return to Top Page]

[Return to IEICE Web Page]

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