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 2021-08-26 10:45
Consideration of primitive action set towards improving activity recognition
Satoshi Nishimura, Shusaku Egami, Ken Fukuda (AIST) PRMU2021-10
Abstract (in Japanese) (See Japanese page) 
(in English) Lots of datasets for activity recognition have been developed. Charades, for instance, provides the term "Throwing a broom somewhere" that includes an object whose state is changed before and after the activity. However, such labeling causes learning the feature of the object rather than the activity itself. Sigurdsson et al, pointed out that the current deep neural network approach can misrecognize the different activities whose object is shared, such as "Holding a broom." We aim to construct a primitive action set as a controlled label set that abstracts the activities from the object and its interpretation. The proposed label set contains 88 terms, and we evaluated its coverage by using three datasets for activity recognition. This paper also provides a plan to evaluate activity recognition accuracy by using the proposed label set.
Keyword (in Japanese) (See Japanese page) 
(in English) Activity recognition / Controlled vocabulary / Ontology / / / / /  
Reference Info. IEICE Tech. Rep., vol. 121, no. 155, PRMU2021-10, pp. 19-23, Aug. 2021.
Paper # PRMU2021-10 
Date of Issue 2021-08-19 (PRMU) 
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 PRMU2021-10

Conference Information
Committee PRMU  
Conference Date 2021-08-26 - 2021-08-26 
Place (in Japanese) (See Japanese page) 
Place (in English) Online 
Topics (in Japanese) (See Japanese page) 
Topics (in English) CV/PR techniques for human-robot cooperation 
Paper Information
Registration To PRMU 
Conference Code 2021-08-PRMU 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Consideration of primitive action set towards improving activity recognition 
Sub Title (in English)  
Keyword(1) Activity recognition  
Keyword(2) Controlled vocabulary  
Keyword(3) Ontology  
Keyword(4)  
Keyword(5)  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Satoshi Nishimura  
1st Author's Affiliation National Institute of Advanced Industrial Science and Technology (AIST) (AIST)
2nd Author's Name Shusaku Egami  
2nd Author's Affiliation National Institute of Advanced Industrial Science and Technology (AIST) (AIST)
3rd Author's Name Ken Fukuda  
3rd Author's Affiliation National Institute of Advanced Industrial Science and Technology (AIST) (AIST)
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 2021-08-26 10:45:00 
Presentation Time 15 minutes 
Registration for PRMU 
Paper # PRMU2021-10 
Volume (vol) vol.121 
Number (no) no.155 
Page pp.19-23 
#Pages
Date of Issue 2021-08-19 (PRMU) 


[Return to Top Page]

[Return to IEICE Web Page]


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