Paper Abstract and Keywords |
Presentation |
2020-09-03 13:10
[Invited Talk]
Manifold modeling in embedded space for image restoration Tatsuya Yokota (Nitech) MI2020-27 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
In this invited talk, I will discuss convolutional neural networks, which have achieved remarkable results in various image processing tasks in recent years. First of all, I would like to introduce some studies that are useful for understanding convolutional neural networks, such as Deep Image Prior and Convolutional Sparse Coding. Next, we explain that convolution operation can be separated into delay embedding and linear transformation. From this point of view, we propose to simplify the convolutional neural network to a combination of delay embedding and auto-encoder, and explain that it can be interpreted as a manifold modeling in embedded space (MMES). We show the similarity between Deep Image Prior and MMES by computational experiments, and argue that perspectives such as shift-invariance and self-similarity are useful for understanding the image defined by the structure of the convolutional neural network. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
Convolutional Neural Networks / Deep Image Prior / Delay-embeddin / Auto-encoder / Shift-invariance / Self-similarity / Manifold modeling in embedded space / |
Reference Info. |
IEICE Tech. Rep., vol. 120, no. 156, MI2020-27, pp. 43-44, Sept. 2020. |
Paper # |
MI2020-27 |
Date of Issue |
2020-08-27 (MI) |
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 |
MI2020-27 |
Conference Information |
Committee |
MI |
Conference Date |
2020-09-03 - 2020-09-03 |
Place (in Japanese) |
(See Japanese page) |
Place (in English) |
Online |
Topics (in Japanese) |
(See Japanese page) |
Topics (in English) |
Medical Image Analysis |
Paper Information |
Registration To |
MI |
Conference Code |
2020-09-MI |
Language |
Japanese |
Title (in Japanese) |
(See Japanese page) |
Sub Title (in Japanese) |
(See Japanese page) |
Title (in English) |
Manifold modeling in embedded space for image restoration |
Sub Title (in English) |
|
Keyword(1) |
Convolutional Neural Networks |
Keyword(2) |
Deep Image Prior |
Keyword(3) |
Delay-embeddin |
Keyword(4) |
Auto-encoder |
Keyword(5) |
Shift-invariance |
Keyword(6) |
Self-similarity |
Keyword(7) |
Manifold modeling in embedded space |
Keyword(8) |
|
1st Author's Name |
Tatsuya Yokota |
1st Author's Affiliation |
Nagoya Institute of Technology (Nitech) |
2nd Author's Name |
|
2nd Author's Affiliation |
() |
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 |
2020-09-03 13:10:00 |
Presentation Time |
45 minutes |
Registration for |
MI |
Paper # |
MI2020-27 |
Volume (vol) |
vol.120 |
Number (no) |
no.156 |
Page |
pp.43-44 |
#Pages |
2 |
Date of Issue |
2020-08-27 (MI) |
|