When Not to Classify: Detection of Reverse Engineering Attacks on DNN Image Classifiers

Yujia Wang; David J. Miller; George Kesidis

doi:10.1109/ICASSP.2019.8682578

When Not to Classify: Detection of Reverse Engineering Attacks on DNN Image Classifiers

Yujia Wang, David J. Miller, George Kesidis

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

3 Scopus citations

Abstract

This paper addresses detection of a reverse engineering (RE) attack targeting a deep neural network (DNN) image classifier; by querying, RE's aim is to discover the classifier's decision rule. RE can enable test-time evasion attacks, which require knowledge of the classifier. Recently, we proposed a quite effective approach (ADA) to detect test-time evasion attacks. In this paper, we extend ADA to detect RE attacks (ADA-RE). We demonstrate our method is successful in detecting »stealthy» RE attacks before they learn enough to launch effective test-time evasion attacks.

Original language	English (US)
Title of host publication	2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	8063-8066
Number of pages	4
ISBN (Electronic)	9781479981311
DOIs	https://doi.org/10.1109/ICASSP.2019.8682578
State	Published - May 2019
Event	44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Brighton, United Kingdom Duration: May 12 2019 → May 17 2019

Publication series

Name	ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Volume	2019-May
ISSN (Print)	1520-6149

Conference

Conference	44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019
Country/Territory	United Kingdom
City	Brighton
Period	5/12/19 → 5/17/19

All Science Journal Classification (ASJC) codes

Software
Signal Processing
Electrical and Electronic Engineering

Access to Document

10.1109/ICASSP.2019.8682578

Cite this

Wang, Y., Miller, D. J., & Kesidis, G. (2019). When Not to Classify: Detection of Reverse Engineering Attacks on DNN Image Classifiers. In 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings (pp. 8063-8066). Article 8682578 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings; Vol. 2019-May). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICASSP.2019.8682578

Wang, Yujia ; Miller, David J. ; Kesidis, George. / When Not to Classify : Detection of Reverse Engineering Attacks on DNN Image Classifiers. 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2019. pp. 8063-8066 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings).

@inproceedings{134870f62e5b474dbe4c482d5c3dac3a,

title = "When Not to Classify: Detection of Reverse Engineering Attacks on DNN Image Classifiers",

abstract = "This paper addresses detection of a reverse engineering (RE) attack targeting a deep neural network (DNN) image classifier; by querying, RE's aim is to discover the classifier's decision rule. RE can enable test-time evasion attacks, which require knowledge of the classifier. Recently, we proposed a quite effective approach (ADA) to detect test-time evasion attacks. In this paper, we extend ADA to detect RE attacks (ADA-RE). We demonstrate our method is successful in detecting »stealthy» RE attacks before they learn enough to launch effective test-time evasion attacks.",

author = "Yujia Wang and Miller, {David J.} and George Kesidis",

note = "Publisher Copyright: {\textcopyright} 2019 IEEE.; 44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 ; Conference date: 12-05-2019 Through 17-05-2019",

year = "2019",

month = may,

doi = "10.1109/ICASSP.2019.8682578",

language = "English (US)",

series = "ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings",

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

pages = "8063--8066",

booktitle = "2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings",

address = "United States",

}

Wang, Y, Miller, DJ & Kesidis, G 2019, When Not to Classify: Detection of Reverse Engineering Attacks on DNN Image Classifiers. in 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings., 8682578, ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, vol. 2019-May, Institute of Electrical and Electronics Engineers Inc., pp. 8063-8066, 44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019, Brighton, United Kingdom, 5/12/19. https://doi.org/10.1109/ICASSP.2019.8682578

When Not to Classify: Detection of Reverse Engineering Attacks on DNN Image Classifiers. / Wang, Yujia; Miller, David J.; Kesidis, George.
2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2019. p. 8063-8066 8682578 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings; Vol. 2019-May).

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

TY - GEN

T1 - When Not to Classify

T2 - 44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019

AU - Wang, Yujia

AU - Miller, David J.

AU - Kesidis, George

PY - 2019/5

Y1 - 2019/5

N2 - This paper addresses detection of a reverse engineering (RE) attack targeting a deep neural network (DNN) image classifier; by querying, RE's aim is to discover the classifier's decision rule. RE can enable test-time evasion attacks, which require knowledge of the classifier. Recently, we proposed a quite effective approach (ADA) to detect test-time evasion attacks. In this paper, we extend ADA to detect RE attacks (ADA-RE). We demonstrate our method is successful in detecting »stealthy» RE attacks before they learn enough to launch effective test-time evasion attacks.

AB - This paper addresses detection of a reverse engineering (RE) attack targeting a deep neural network (DNN) image classifier; by querying, RE's aim is to discover the classifier's decision rule. RE can enable test-time evasion attacks, which require knowledge of the classifier. Recently, we proposed a quite effective approach (ADA) to detect test-time evasion attacks. In this paper, we extend ADA to detect RE attacks (ADA-RE). We demonstrate our method is successful in detecting »stealthy» RE attacks before they learn enough to launch effective test-time evasion attacks.

UR - http://www.scopus.com/inward/record.url?scp=85068969241&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85068969241&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2019.8682578

DO - 10.1109/ICASSP.2019.8682578

M3 - Conference contribution

AN - SCOPUS:85068969241

T3 - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

SP - 8063

EP - 8066

BT - 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 12 May 2019 through 17 May 2019

ER -

Wang Y, Miller DJ , Kesidis G. When Not to Classify: Detection of Reverse Engineering Attacks on DNN Image Classifiers. In 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2019. p. 8063-8066. 8682578. (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings). doi: 10.1109/ICASSP.2019.8682578

When Not to Classify: Detection of Reverse Engineering Attacks on DNN Image Classifiers

Abstract

Publication series

Conference

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this