Presentation 2024-03-22
Adversarial Examples with Missing Perturbation Using Laser Irradiation
Daisuke Kosuge, Hayato Watanabe, Taiga Manabe, Yoshihisa Takayama, Toshihiro Ohigashi,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In recent years, neural networks have made remarkable progress in the field of image processing and other areas, and the development of technologies that will provide a new foundation for our daily lives is underway. However, there exist hostile samples that intentionally misclassify images to a different class by exploiting vulnerabilities in image processing. We propose a method of attack by Adversarial Examples that uses lasers to facilitate control of misclassification. Studies have already shown that road signs can be misclassified by irradiating them with lasers. We address the question of whether an attack can be more effective when the attacker has the advantage of having postings available. Instead of simply irradiating lasers, we prepare Adversarial Examples with missing perturbation added to the input image to facilitate manipulation of features and make the attack easier. Furthermore, experiments were conducted to irradiate a Adversarial Examples with a missing perturbation, and the effectiveness of the missing perturbation was demonstrated.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Missing Perturbation / Neural Network / Adversarial Examples / Laser
Paper # ICSS2023-97
Date of Issue 2024-03-14 (ICSS)

Conference Information
Committee ICSS / IPSJ-SPT
Conference Date 2024/3/21(2days)
Place (in Japanese) (See Japanese page)
Place (in English) OIST
Topics (in Japanese) (See Japanese page)
Topics (in English) Security, Trust, etc.
Chair Daisuke Inoue(NICT)
Vice Chair Akira Yamada(Kobe Univ.) / Toshihiro Yamauchi(Okayama Univ.)
Secretary Akira Yamada(Mitsubishi Electric) / Toshihiro Yamauchi(Univ. of Electro-Comm.)
Assistant Yo Kanemoto(NTT) / Masaya Sato(Okayama Prefectural Univ.)

Paper Information
Registration To Technical Committee on Information and Communication System Security / Special Interest Group on Security Psychology and Trust
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Adversarial Examples with Missing Perturbation Using Laser Irradiation
Sub Title (in English)
Keyword(1) Missing Perturbation
Keyword(2) Neural Network
Keyword(3) Adversarial Examples
Keyword(4) Laser
1st Author's Name Daisuke Kosuge
1st Author's Affiliation Tokai University(Tokai Univ.)
2nd Author's Name Hayato Watanabe
2nd Author's Affiliation Tokai University(Tokai Univ.)
3rd Author's Name Taiga Manabe
3rd Author's Affiliation Tokai University(Tokai Univ.)
4th Author's Name Yoshihisa Takayama
4th Author's Affiliation Tokai University(Tokai Univ.)
5th Author's Name Toshihiro Ohigashi
5th Author's Affiliation Tokai University(Tokai Univ.)
Date 2024-03-22
Paper # ICSS2023-97
Volume (vol) vol.123
Number (no) ICSS-448
Page pp.pp.201-207(ICSS),
#Pages 7
Date of Issue 2024-03-14 (ICSS)