Paper Abstract and Keywords |
Presentation |
2019-01-23 15:40
Measuring the Convolution Neural Network similarities trained with different dataset using SVCCA Toya Teramoto, Hayaru Shouno (UEC) NC2018-40 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
Deep Convolutional Neural Network (DCNN) is a successful model in the field of computer vision such like image classification and segmentation and so on.
However, the DCNN has still been regarded as block box.
And many researches are try to analyze what kind of feature expression DCNN learns in recent years.
Understanding the DCNN is important because it leads to an improvement in performance, and it is also available in products where an black-box system can not be used.
In this research, in oder to analyzing, we investigate the difference of feature representations in DCNNs which are trained with similar datasets.
We also investigate the difference from the viewpoint of the model structures.
In this study, we compare models learned with different but similar datasets and consider how much difference will occur in the network by changing the capacity of the model.
As a results, we obtain the network capacity is important to keep similarity in representations. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
Deep Learning / Neocognitron / Canonical Correlation Analysis / / / / / |
Reference Info. |
IEICE Tech. Rep., vol. 118, no. 414, NC2018-40, pp. 11-16, Jan. 2019. |
Paper # |
NC2018-40 |
Date of Issue |
2019-01-16 (NC) |
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 |
NC2018-40 |
Conference Information |
Committee |
NLP NC |
Conference Date |
2019-01-23 - 2019-01-24 |
Place (in Japanese) |
(See Japanese page) |
Place (in English) |
The Centennial Hall, Hokkaido Univ. |
Topics (in Japanese) |
(See Japanese page) |
Topics (in English) |
General Implementation of Neuro Computing, Analysis and Modeling of Human Science, etc. |
Paper Information |
Registration To |
NC |
Conference Code |
2019-01-NLP-NC |
Language |
Japanese |
Title (in Japanese) |
(See Japanese page) |
Sub Title (in Japanese) |
(See Japanese page) |
Title (in English) |
Measuring the Convolution Neural Network similarities trained with different dataset using SVCCA |
Sub Title (in English) |
|
Keyword(1) |
Deep Learning |
Keyword(2) |
Neocognitron |
Keyword(3) |
Canonical Correlation Analysis |
Keyword(4) |
|
Keyword(5) |
|
Keyword(6) |
|
Keyword(7) |
|
Keyword(8) |
|
1st Author's Name |
Toya Teramoto |
1st Author's Affiliation |
University of Electro-Communications (UEC) |
2nd Author's Name |
Hayaru Shouno |
2nd Author's Affiliation |
University of Electro-Communications (UEC) |
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 |
2019-01-23 15:40:00 |
Presentation Time |
20 minutes |
Registration for |
NC |
Paper # |
NC2018-40 |
Volume (vol) |
vol.118 |
Number (no) |
no.414 |
Page |
pp.11-16 |
#Pages |
6 |
Date of Issue |
2019-01-16 (NC) |
|