Paper Abstract and Keywords |
Presentation |
2024-03-04 09:12
Creating Adversarial Examples to Deceive Both Humans and Machine Learning Models Ko Fujimori (Waseda Univ.), Toshiki Shibahara (NTT), Daiki Chiba (NTT Security), Mitsuaki Akiyama (NTT), Masato Uchida (Waseda Univ.) PRMU2023-65 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
One of the vulnerability attacks against neural networks is the generation of Adversarial Examples (AE), which induce misclassification by adding minimal noise to input data.
The ``attack success'' by AE is defined as causing a machine learning model to misclassify without the noise being recognized by humans.
However, existing research on AE attack methods often focuses solely on causing misclassification of machine learning models and may not evaluate the visibility of the noise.
Evaluation experiments conducted in the same conditions as the paper that proposed the prominent attack method, the Fast Gradient Sign Method, have confirmed that the majority of AEs are perceptible with noise.
Therefore, in this study, we propose a method for creating AEs that appear visually natural. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
Adversarial Example / Fast Gradient Sign Method / Noise Visibility / User Surveys / / / / |
Reference Info. |
IEICE Tech. Rep., vol. 123, no. 409, PRMU2023-65, pp. 82-87, March 2024. |
Paper # |
PRMU2023-65 |
Date of Issue |
2024-02-25 (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 |
PRMU2023-65 |
Conference Information |
Committee |
PRMU IBISML IPSJ-CVIM |
Conference Date |
2024-03-03 - 2024-03-04 |
Place (in Japanese) |
(See Japanese page) |
Place (in English) |
Hiroshima Univ. Higashi-Hiroshima campus |
Topics (in Japanese) |
(See Japanese page) |
Topics (in English) |
|
Paper Information |
Registration To |
PRMU |
Conference Code |
2024-03-PRMU-IBISML-CVIM |
Language |
Japanese |
Title (in Japanese) |
(See Japanese page) |
Sub Title (in Japanese) |
(See Japanese page) |
Title (in English) |
Creating Adversarial Examples to Deceive Both Humans and Machine Learning Models |
Sub Title (in English) |
|
Keyword(1) |
Adversarial Example |
Keyword(2) |
Fast Gradient Sign Method |
Keyword(3) |
Noise Visibility |
Keyword(4) |
User Surveys |
Keyword(5) |
|
Keyword(6) |
|
Keyword(7) |
|
Keyword(8) |
|
1st Author's Name |
Ko Fujimori |
1st Author's Affiliation |
Waseda University (Waseda Univ.) |
2nd Author's Name |
Toshiki Shibahara |
2nd Author's Affiliation |
Nippon Telegraph And Telephone Corporation (NTT) |
3rd Author's Name |
Daiki Chiba |
3rd Author's Affiliation |
NTT Security Holdings (NTT Security) |
4th Author's Name |
Mitsuaki Akiyama |
4th Author's Affiliation |
Nippon Telegraph And Telephone Corporation (NTT) |
5th Author's Name |
Masato Uchida |
5th Author's Affiliation |
Waseda University (Waseda Univ.) |
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 |
2024-03-04 09:12:00 |
Presentation Time |
12 minutes |
Registration for |
PRMU |
Paper # |
PRMU2023-65 |
Volume (vol) |
vol.123 |
Number (no) |
no.409 |
Page |
pp.82-87 |
#Pages |
6 |
Date of Issue |
2024-02-25 (PRMU) |
|