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 2022-10-21 15:55
Fine-grained Action Recognition Using Hand Trajectory Features for First-Person Cooking Videos
Sota Miyamoto (Tokyo Tech), Yagi Takuma (UTokyo), Ushiku Yoshitaka, Atsushi Hashimoto (OSX), Nakamasa Inoue (Tokyo Tech) PRMU2022-29
Abstract (in Japanese) (See Japanese page) 
(in English) Action recognition in kitchen environments is a crucial topic in computer vision and robotics
with application toward fully automatic cooking. In this paper, we propose a method of fine-grained action recognition using
hand trajectory features. Our method extracts seven low-level features of hand trajectories and feeds them into a Transformer to classify fine-grained actions. In experiments, we annotated fine-grained labels of cutting
methods such as slice and mince on EPIC-KITCHENS and Ego4D datasets for evaluating our method. We showed that hand
trajectory features improve the recognition accuracy over the baseline method using RGB, optical flow, and audio features.
Keyword (in Japanese) (See Japanese page) 
(in English) Fine-grained action recognition / Cooking videos / Hand-trajectory features / Multimodal deep learning / / / /  
Reference Info. IEICE Tech. Rep., vol. 122, no. 223, PRMU2022-29, pp. 41-46, Oct. 2022.
Paper # PRMU2022-29 
Date of Issue 2022-10-14 (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 PRMU2022-29

Conference Information
Committee PRMU  
Conference Date 2022-10-21 - 2022-10-22 
Place (in Japanese) (See Japanese page) 
Place (in English) Miraikan - The National Museum of Emerging Science and Innovation 
Topics (in Japanese) (See Japanese page) 
Topics (in English) Recognition and understanding related to people 
Paper Information
Registration To PRMU 
Conference Code 2022-10-PRMU 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Fine-grained Action Recognition Using Hand Trajectory Features for First-Person Cooking Videos 
Sub Title (in English)  
Keyword(1) Fine-grained action recognition  
Keyword(2) Cooking videos  
Keyword(3) Hand-trajectory features  
Keyword(4) Multimodal deep learning  
1st Author's Name Sota Miyamoto  
1st Author's Affiliation Tokyo Institute of Technology (Tokyo Tech)
2nd Author's Name Yagi Takuma  
2nd Author's Affiliation The University of Tokyo (UTokyo)
3rd Author's Name Ushiku Yoshitaka  
3rd Author's Affiliation OMRON SINIC X (OSX)
4th Author's Name Atsushi Hashimoto  
4th Author's Affiliation OMRON SINIC X (OSX)
5th Author's Name Nakamasa Inoue  
5th Author's Affiliation Tokyo Institute of Technology (Tokyo Tech)
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 2022-10-21 15:55:00 
Presentation Time 15 minutes 
Registration for PRMU 
Paper # PRMU2022-29 
Volume (vol) vol.122 
Number (no) no.223 
Page pp.41-46 
Date of Issue 2022-10-14 (PRMU) 

[Return to Top Page]

[Return to IEICE Web Page]

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