Adversarial Robustness of Flow-based Image Super-Resolution

Junha Park; Jun Ho Choi; Jong Seok Lee

doi:10.1109/MMSP55362.2022.9949274

Adversarial Robustness of Flow-based Image Super-Resolution

Junha Park, Jun Ho Choi, Jong Seok Lee

School of Integrated Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

This paper investigates the robustness of deep image super-resolution models using normalizing flow against adversarial attacks. Attack methods specific to flow-based super-resolution models are formulated, and the performance and influences of the attacks are analyzed. We show that flow-based super-resolution models are highly vulnerable to attacks, which are even more serious than other super-resolution models. Potential remedies to the vulnerability are also evaluated.

Original language	English
Title of host publication	2022 IEEE 24th International Workshop on Multimedia Signal Processing, MMSP 2022
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781665471893
DOIs	https://doi.org/10.1109/MMSP55362.2022.9949274
Publication status	Published - 2022
Event	24th IEEE International Workshop on Multimedia Signal Processing, MMSP 2022 - Shanghai, China Duration: 2022 Sept 26 → 2022 Sept 28

Publication series

Name	2022 IEEE 24th International Workshop on Multimedia Signal Processing, MMSP 2022

Conference

Conference	24th IEEE International Workshop on Multimedia Signal Processing, MMSP 2022
Country/Territory	China
City	Shanghai
Period	22/9/26 → 22/9/28

Bibliographical note

Publisher Copyright:
© 2022 IEEE.

All Science Journal Classification (ASJC) codes

Computer Vision and Pattern Recognition
Signal Processing
Media Technology

Access to Document

10.1109/MMSP55362.2022.9949274

Cite this

Park, J., Choi, J. H., & Lee, J. S. (2022). Adversarial Robustness of Flow-based Image Super-Resolution. In 2022 IEEE 24th International Workshop on Multimedia Signal Processing, MMSP 2022 (2022 IEEE 24th International Workshop on Multimedia Signal Processing, MMSP 2022). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/MMSP55362.2022.9949274

@inproceedings{f48386a3e6684c608554beec85a60023,

title = "Adversarial Robustness of Flow-based Image Super-Resolution",

abstract = "This paper investigates the robustness of deep image super-resolution models using normalizing flow against adversarial attacks. Attack methods specific to flow-based super-resolution models are formulated, and the performance and influences of the attacks are analyzed. We show that flow-based super-resolution models are highly vulnerable to attacks, which are even more serious than other super-resolution models. Potential remedies to the vulnerability are also evaluated.",

keywords = "Adversarial attack, defense, normalizing flow, super-resolution",

author = "Junha Park and Choi, {Jun Ho} and Lee, {Jong Seok}",

note = "Publisher Copyright: {\textcopyright} 2022 IEEE.; 24th IEEE International Workshop on Multimedia Signal Processing, MMSP 2022 ; Conference date: 26-09-2022 Through 28-09-2022",

year = "2022",

doi = "10.1109/MMSP55362.2022.9949274",

language = "English",

series = "2022 IEEE 24th International Workshop on Multimedia Signal Processing, MMSP 2022",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2022 IEEE 24th International Workshop on Multimedia Signal Processing, MMSP 2022",

address = "United States",

}

Park, J, Choi, JH & Lee, JS 2022, Adversarial Robustness of Flow-based Image Super-Resolution. in 2022 IEEE 24th International Workshop on Multimedia Signal Processing, MMSP 2022. 2022 IEEE 24th International Workshop on Multimedia Signal Processing, MMSP 2022, Institute of Electrical and Electronics Engineers Inc., 24th IEEE International Workshop on Multimedia Signal Processing, MMSP 2022, Shanghai, China, 22/9/26. https://doi.org/10.1109/MMSP55362.2022.9949274

Adversarial Robustness of Flow-based Image Super-Resolution. / Park, Junha; Choi, Jun Ho; Lee, Jong Seok.
2022 IEEE 24th International Workshop on Multimedia Signal Processing, MMSP 2022. Institute of Electrical and Electronics Engineers Inc., 2022. (2022 IEEE 24th International Workshop on Multimedia Signal Processing, MMSP 2022).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Adversarial Robustness of Flow-based Image Super-Resolution

AU - Park, Junha

AU - Choi, Jun Ho

AU - Lee, Jong Seok

PY - 2022

Y1 - 2022

N2 - This paper investigates the robustness of deep image super-resolution models using normalizing flow against adversarial attacks. Attack methods specific to flow-based super-resolution models are formulated, and the performance and influences of the attacks are analyzed. We show that flow-based super-resolution models are highly vulnerable to attacks, which are even more serious than other super-resolution models. Potential remedies to the vulnerability are also evaluated.

AB - This paper investigates the robustness of deep image super-resolution models using normalizing flow against adversarial attacks. Attack methods specific to flow-based super-resolution models are formulated, and the performance and influences of the attacks are analyzed. We show that flow-based super-resolution models are highly vulnerable to attacks, which are even more serious than other super-resolution models. Potential remedies to the vulnerability are also evaluated.

KW - Adversarial attack

KW - defense

KW - normalizing flow

KW - super-resolution

UR - http://www.scopus.com/inward/record.url?scp=85143609341&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85143609341&partnerID=8YFLogxK

U2 - 10.1109/MMSP55362.2022.9949274

DO - 10.1109/MMSP55362.2022.9949274

M3 - Conference contribution

AN - SCOPUS:85143609341

T3 - 2022 IEEE 24th International Workshop on Multimedia Signal Processing, MMSP 2022

BT - 2022 IEEE 24th International Workshop on Multimedia Signal Processing, MMSP 2022

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 24th IEEE International Workshop on Multimedia Signal Processing, MMSP 2022

Y2 - 26 September 2022 through 28 September 2022

ER -

Adversarial Robustness of Flow-based Image Super-Resolution

Abstract

Publication series

Conference

Bibliographical note

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this