Modeling Adversarial Noise for Adversarial Training

Dawei Zhou; Nannan Wang; Bo Han; Tongliang Liu

Modeling Adversarial Noise for Adversarial Training

Dawei Zhou, Nannan Wang, Bo Han, Tongliang Liu^*

^*Corresponding author for this work

Department of Computer Science

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

6 Citations (Scopus)

Abstract

Deep neural networks have been demonstrated to be vulnerable to adversarial noise, promoting the development of defense against adversarial attacks. Motivated by the fact that adversarial noise contains well-generalizing features and that the relationship between adversarial data and natural data can help infer natural data and make reliable predictions, in this paper, we study to model adversarial noise by learning the transition relationship between adversarial labels (i.e. the flipped labels used to generate adversarial data) and natural labels (i.e. the ground truth labels of the natural data). Specifically, we introduce an instance-dependent transition matrix to relate adversarial labels and natural labels, which can be seamlessly embedded with the target model (enabling us to model stronger adaptive adversarial noise). Empirical evaluations demonstrate that our method could effectively improve adversarial accuracy.

Original language	English
Title of host publication	Proceedings of 39th International Conference on Machine Learning (ICML 2022)
Editors	Kamalika Chaudhuri, Stefanie Jegelka, Le Song, Csaba Szepesvari, Gang Niu, Sivan Sabato
Publisher	ML Research Press
Pages	27353-27366
Number of pages	14
Publication status	Published - 17 Jul 2022
Event	39th International Conference on Machine Learning, ICML 2022 - Baltimore Convention Center , Baltimore, Maryland, United States Duration: 17 Jul 2022 → 23 Jul 2022 https://icml.cc/Conferences/2022

Publication series

Name	Proceedings of Machine Learning Research
Volume	162
ISSN (Print)	2640-3498

Conference

Conference	39th International Conference on Machine Learning, ICML 2022
Country/Territory	United States
City	Baltimore, Maryland
Period	17/07/22 → 23/07/22
Internet address	https://icml.cc/Conferences/2022

Access to Document

https://proceedings.mlr.press/v162/zhou22k/zhou22k.pdf

Cite this

Zhou, D., Wang, N., Han, B., & Liu, T. (2022). Modeling Adversarial Noise for Adversarial Training. In K. Chaudhuri, S. Jegelka, L. Song, C. Szepesvari, G. Niu, & S. Sabato (Eds.), Proceedings of 39th International Conference on Machine Learning (ICML 2022) (pp. 27353-27366). (Proceedings of Machine Learning Research; Vol. 162). ML Research Press. https://proceedings.mlr.press/v162/zhou22k/zhou22k.pdf

@inproceedings{54dba9df1b8448b0b71acf21939e0439,

title = "Modeling Adversarial Noise for Adversarial Training",

abstract = "Deep neural networks have been demonstrated to be vulnerable to adversarial noise, promoting the development of defense against adversarial attacks. Motivated by the fact that adversarial noise contains well-generalizing features and that the relationship between adversarial data and natural data can help infer natural data and make reliable predictions, in this paper, we study to model adversarial noise by learning the transition relationship between adversarial labels (i.e. the flipped labels used to generate adversarial data) and natural labels (i.e. the ground truth labels of the natural data). Specifically, we introduce an instance-dependent transition matrix to relate adversarial labels and natural labels, which can be seamlessly embedded with the target model (enabling us to model stronger adaptive adversarial noise). Empirical evaluations demonstrate that our method could effectively improve adversarial accuracy.",

author = "Dawei Zhou and Nannan Wang and Bo Han and Tongliang Liu",

note = "This work was supported in part by the National Key Research and Development Program of China under Grant 2018AAA0103202, in part by the National Natural Science Foundation of China under Grant 61922066, 61876142, 62036007 and 62006202, in part by the Technology Innovation Leading Program of Shaanxi under Grant 2022QFY01-15, in part by Open Research Projects of Zhejiang Lab under Grant 2021KG0AB01, in part by the RGC Early Career Scheme No. 22200720, in part by Guangdong Basic and Applied Basic Research Foundation No. 2022A1515011652, in part by Australian Research Council Projects DE190101473, IC-190100031, and DP-220102121, in part by the Fundamental Research Funds for the Central Universities, and in part by the Innovation Fund of Xidian University. The authors thank the reviewers and the meta-reviewer for their helpful and constructive comments on this work. Thanks to Chaojian Yu for his important advice on Section attack the transition matrix. ; 39th International Conference on Machine Learning, ICML 2022 ; Conference date: 17-07-2022 Through 23-07-2022",

year = "2022",

month = jul,

day = "17",

language = "English",

series = "Proceedings of Machine Learning Research",

publisher = "ML Research Press",

pages = "27353--27366",

editor = "Kamalika Chaudhuri and Stefanie Jegelka and Le Song and Csaba Szepesvari and Gang Niu and Sivan Sabato",

booktitle = "Proceedings of 39th International Conference on Machine Learning (ICML 2022)",

url = "https://icml.cc/Conferences/2022",

}

Zhou, D, Wang, N, Han, B & Liu, T 2022, Modeling Adversarial Noise for Adversarial Training. in K Chaudhuri, S Jegelka, L Song, C Szepesvari, G Niu & S Sabato (eds), Proceedings of 39th International Conference on Machine Learning (ICML 2022). Proceedings of Machine Learning Research, vol. 162, ML Research Press, pp. 27353-27366, 39th International Conference on Machine Learning, ICML 2022, Baltimore, Maryland, United States, 17/07/22. <https://proceedings.mlr.press/v162/zhou22k/zhou22k.pdf>

Modeling Adversarial Noise for Adversarial Training. / Zhou, Dawei; Wang, Nannan; Han, Bo et al.
Proceedings of 39th International Conference on Machine Learning (ICML 2022). ed. / Kamalika Chaudhuri; Stefanie Jegelka; Le Song; Csaba Szepesvari; Gang Niu; Sivan Sabato. ML Research Press, 2022. p. 27353-27366 (Proceedings of Machine Learning Research; Vol. 162).

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

TY - GEN

T1 - Modeling Adversarial Noise for Adversarial Training

AU - Zhou, Dawei

AU - Wang, Nannan

AU - Han, Bo

AU - Liu, Tongliang

N1 - This work was supported in part by the National Key Research and Development Program of China under Grant 2018AAA0103202, in part by the National Natural Science Foundation of China under Grant 61922066, 61876142, 62036007 and 62006202, in part by the Technology Innovation Leading Program of Shaanxi under Grant 2022QFY01-15, in part by Open Research Projects of Zhejiang Lab under Grant 2021KG0AB01, in part by the RGC Early Career Scheme No. 22200720, in part by Guangdong Basic and Applied Basic Research Foundation No. 2022A1515011652, in part by Australian Research Council Projects DE190101473, IC-190100031, and DP-220102121, in part by the Fundamental Research Funds for the Central Universities, and in part by the Innovation Fund of Xidian University. The authors thank the reviewers and the meta-reviewer for their helpful and constructive comments on this work. Thanks to Chaojian Yu for his important advice on Section attack the transition matrix.

PY - 2022/7/17

Y1 - 2022/7/17

N2 - Deep neural networks have been demonstrated to be vulnerable to adversarial noise, promoting the development of defense against adversarial attacks. Motivated by the fact that adversarial noise contains well-generalizing features and that the relationship between adversarial data and natural data can help infer natural data and make reliable predictions, in this paper, we study to model adversarial noise by learning the transition relationship between adversarial labels (i.e. the flipped labels used to generate adversarial data) and natural labels (i.e. the ground truth labels of the natural data). Specifically, we introduce an instance-dependent transition matrix to relate adversarial labels and natural labels, which can be seamlessly embedded with the target model (enabling us to model stronger adaptive adversarial noise). Empirical evaluations demonstrate that our method could effectively improve adversarial accuracy.

AB - Deep neural networks have been demonstrated to be vulnerable to adversarial noise, promoting the development of defense against adversarial attacks. Motivated by the fact that adversarial noise contains well-generalizing features and that the relationship between adversarial data and natural data can help infer natural data and make reliable predictions, in this paper, we study to model adversarial noise by learning the transition relationship between adversarial labels (i.e. the flipped labels used to generate adversarial data) and natural labels (i.e. the ground truth labels of the natural data). Specifically, we introduce an instance-dependent transition matrix to relate adversarial labels and natural labels, which can be seamlessly embedded with the target model (enabling us to model stronger adaptive adversarial noise). Empirical evaluations demonstrate that our method could effectively improve adversarial accuracy.

UR - https://proceedings.mlr.press/v162/zhou22k.html

UR - https://proceedings.mlr.press/v162/

M3 - Conference proceeding

T3 - Proceedings of Machine Learning Research

SP - 27353

EP - 27366

BT - Proceedings of 39th International Conference on Machine Learning (ICML 2022)

A2 - Chaudhuri, Kamalika

A2 - Jegelka, Stefanie

A2 - Song, Le

A2 - Szepesvari, Csaba

A2 - Niu, Gang

A2 - Sabato, Sivan

PB - ML Research Press

T2 - 39th International Conference on Machine Learning, ICML 2022

Y2 - 17 July 2022 through 23 July 2022

ER -

Modeling Adversarial Noise for Adversarial Training

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this