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 2022-05-17 15:05
Generating patch-wise adversarial examples for avoidance of face recognition system and verification of its robustness
Hiroto Takiwaki, Minoru Kuribayashi, Nobuo Funabiki (Okayama univ.) IT2022-5 EMM2022-5
Abstract (in Japanese) (See Japanese page) 
(in English) Advances in machine learning technologies such as Convolutional Neural Networks (CNN) have made it possible to identify individuals by face recognition. The system can be constructed to detect the presence of a specific type of object. However, if these technologies are abused, individuals can be automatically tracked and their privacy will be violated. Therefore, it is necessary to develop a technique for avoiding automated human tracking systems based on a face identification technique. Altough it has been considered to generate an adversarial noise to avoid automated tracking, changing position and size of the generated noise has not been examined in detail. In this paper, we use a small and transparent patch to generate adversarial examples by adding noise at various positions and sizes, and evaluated its performance. In addition, the robustness of the system against disturbance is considered by shifting the position of the attachment. The experimental results show that adversarial exanples can be generated for various parts of the face. The attack success rate is high when the noise is close to the facial features such as eyes, nose, and chin.
Keyword (in Japanese) (See Japanese page) 
(in English) Face recognition / Adversarial example / FGSM / robustness / / / /  
Reference Info. IEICE Tech. Rep., vol. 122, no. 26, EMM2022-5, pp. 23-28, May 2022.
Paper # EMM2022-5 
Date of Issue 2022-05-10 (IT, EMM) 
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 IT2022-5 EMM2022-5

Conference Information
Committee IT EMM  
Conference Date 2022-05-17 - 2022-05-18 
Place (in Japanese) (See Japanese page) 
Place (in English) Gifu University 
Topics (in Japanese) (See Japanese page) 
Topics (in English) Information Security, Information Theory, Information Hiding, etc. 
Paper Information
Registration To EMM 
Conference Code 2022-05-IT-EMM 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Generating patch-wise adversarial examples for avoidance of face recognition system and verification of its robustness 
Sub Title (in English)  
Keyword(1) Face recognition  
Keyword(2) Adversarial example  
Keyword(3) FGSM  
Keyword(4) robustness  
Keyword(5)  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Hiroto Takiwaki  
1st Author's Affiliation Okayama University (Okayama univ.)
2nd Author's Name Minoru Kuribayashi  
2nd Author's Affiliation Okayama University (Okayama univ.)
3rd Author's Name Nobuo Funabiki  
3rd Author's Affiliation Okayama University (Okayama univ.)
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 2022-05-17 15:05:00 
Presentation Time 25 minutes 
Registration for EMM 
Paper # IT2022-5, EMM2022-5 
Volume (vol) vol.122 
Number (no) no.25(IT), no.26(EMM) 
Page pp.23-28 
#Pages
Date of Issue 2022-05-10 (IT, EMM) 


[Return to Top Page]

[Return to IEICE Web Page]


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