Paper Abstract and Keywords |
Presentation |
2017-02-18 14:10
Fine-grained Pedestrian Classification by Fusing Multiple CNN Models Yoshihito Kokubo, Yu Wang, Jien Kato, Kenji Mase (Nagoya Univ.) PRMU2016-166 CNR2016-33 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
Fine-grained categories of pedestrians lead to higher level understanding on pedestrians as well as their activities, therefore is a perspective key component in the next generation smart vehicle and surveillance systems. In this study, we propose fine-grained pedestrian classification method by integrating features obtained from partial image and whole image using Convolutional Neural Network (CNN), which has dramatically improved results in various fields in recent years. Generally, when the partial image is given as an input to the network, detailed features can be expressed for the local portion, but the entire information is missing. On the other hand, when the entire image is given as an input to the network, it is possible to consider the appearance of the entire image, but detailed features can not be acquired compared with learning using a partial image. In the proposed method, a CNN model is learned for each patch of partial images and an entire image. Next, by re-training the fusion network in which all CNN models are integrated, the proposed method can eliminate mutual disadvantages and implement more accurate classification. In the experiment, we confirmed that the proposed method exceeds the baseline accuracy. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
image classification / deep learning / / / / / / |
Reference Info. |
IEICE Tech. Rep., vol. 116, no. 461, PRMU2016-166, pp. 81-85, Feb. 2017. |
Paper # |
PRMU2016-166 |
Date of Issue |
2017-02-11 (PRMU, CNR) |
ISSN |
Print edition: ISSN 0913-5685 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 |
PRMU2016-166 CNR2016-33 |
Conference Information |
Committee |
PRMU CNR |
Conference Date |
2017-02-18 - 2017-02-19 |
Place (in Japanese) |
(See Japanese page) |
Place (in English) |
|
Topics (in Japanese) |
(See Japanese page) |
Topics (in English) |
|
Paper Information |
Registration To |
PRMU |
Conference Code |
2017-02-PRMU-CNR |
Language |
Japanese |
Title (in Japanese) |
(See Japanese page) |
Sub Title (in Japanese) |
(See Japanese page) |
Title (in English) |
Fine-grained Pedestrian Classification by Fusing Multiple CNN Models |
Sub Title (in English) |
|
Keyword(1) |
image classification |
Keyword(2) |
deep learning |
Keyword(3) |
|
Keyword(4) |
|
Keyword(5) |
|
Keyword(6) |
|
Keyword(7) |
|
Keyword(8) |
|
1st Author's Name |
Yoshihito Kokubo |
1st Author's Affiliation |
Nagoya University (Nagoya Univ.) |
2nd Author's Name |
Yu Wang |
2nd Author's Affiliation |
Nagoya University (Nagoya Univ.) |
3rd Author's Name |
Jien Kato |
3rd Author's Affiliation |
Nagoya University (Nagoya Univ.) |
4th Author's Name |
Kenji Mase |
4th Author's Affiliation |
Nagoya University (Nagoya 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 |
2017-02-18 14:10:00 |
Presentation Time |
25 minutes |
Registration for |
PRMU |
Paper # |
PRMU2016-166, CNR2016-33 |
Volume (vol) |
vol.116 |
Number (no) |
no.461(PRMU), no.462(CNR) |
Page |
pp.81-85 |
#Pages |
5 |
Date of Issue |
2017-02-11 (PRMU, CNR) |
|