Dynamic security-level maximization for stabilized parallel deep learning architectures in surveillance applications

Joongheon Kim; Yeong Jong Mo; Woojoo Lee; Daehun Nyang

doi:10.1109/PAC.2017.22

Dynamic security-level maximization for stabilized parallel deep learning architectures in surveillance applications

Joongheon Kim, Yeong Jong Mo, Woojoo Lee, Daehun Nyang

Department of Cyber Security

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

12 Scopus citations

Abstract

This paper introduces a new surveillance platform which is equipped with multiple parallel deep learning frameworks. The deep learning frameworks are used for the face recognition of input image and video streams from CCTV cameras in security applications. Each deep learning framework has its own accuracy (related to recognition performance) and operation time (related to system stability) those are in tradeoff relationship. Based on this system architecture, a new dynamic control algorithm which selects one deep learning framework for time-average security-level (i.e., machine learning accuracy for recognition and classification) maximization under the consideration of system stability. The performance of the proposed algorithm was evaluated and also verified that it achieves desired performance.

Original language	English
Title of host publication	Proceedings - 2017 IEEE Symposium on Privacy-Aware Computing, PAC 2017
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	192-193
Number of pages	2
ISBN (Electronic)	9781538610275
DOIs	https://doi.org/10.1109/PAC.2017.22
State	Published - 4 Dec 2017
Event	1st IEEE Symposium on Privacy-Aware Computing, PAC 2017 - Washington, United States Duration: 1 Aug 2017 → 3 Aug 2017

Publication series

Name	Proceedings - 2017 IEEE Symposium on Privacy-Aware Computing, PAC 2017
Volume	2017-January

Conference

Conference	1st IEEE Symposium on Privacy-Aware Computing, PAC 2017
Country/Territory	United States
City	Washington
Period	1/08/17 → 3/08/17

Bibliographical note

Funding Information:
ACKNOWLEDGEMENT This work was supported by National Research Foundation of Korea (No: 2016R1C1B1015406).

Publisher Copyright:
© 2017 IEEE.

Keywords

Deep Learning
Lyapunov Optimization
Security

Access to Document

10.1109/PAC.2017.22

Cite this

Kim, J., Mo, Y. J., Lee, W., & Nyang, D. (2017). Dynamic security-level maximization for stabilized parallel deep learning architectures in surveillance applications. In Proceedings - 2017 IEEE Symposium on Privacy-Aware Computing, PAC 2017 (pp. 192-193). (Proceedings - 2017 IEEE Symposium on Privacy-Aware Computing, PAC 2017; Vol. 2017-January). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/PAC.2017.22

Kim, Joongheon ; Mo, Yeong Jong ; Lee, Woojoo et al. / Dynamic security-level maximization for stabilized parallel deep learning architectures in surveillance applications. Proceedings - 2017 IEEE Symposium on Privacy-Aware Computing, PAC 2017. Institute of Electrical and Electronics Engineers Inc., 2017. pp. 192-193 (Proceedings - 2017 IEEE Symposium on Privacy-Aware Computing, PAC 2017).

@inproceedings{bca0cd2c121d4b88ad3a1b902da74cf6,

title = "Dynamic security-level maximization for stabilized parallel deep learning architectures in surveillance applications",

abstract = "This paper introduces a new surveillance platform which is equipped with multiple parallel deep learning frameworks. The deep learning frameworks are used for the face recognition of input image and video streams from CCTV cameras in security applications. Each deep learning framework has its own accuracy (related to recognition performance) and operation time (related to system stability) those are in tradeoff relationship. Based on this system architecture, a new dynamic control algorithm which selects one deep learning framework for time-average security-level (i.e., machine learning accuracy for recognition and classification) maximization under the consideration of system stability. The performance of the proposed algorithm was evaluated and also verified that it achieves desired performance.",

keywords = "Deep Learning, Lyapunov Optimization, Security",

author = "Joongheon Kim and Mo, {Yeong Jong} and Woojoo Lee and Daehun Nyang",

note = "Funding Information: ACKNOWLEDGEMENT This work was supported by National Research Foundation of Korea (No: 2016R1C1B1015406). Publisher Copyright: {\textcopyright} 2017 IEEE.; 1st IEEE Symposium on Privacy-Aware Computing, PAC 2017 ; Conference date: 01-08-2017 Through 03-08-2017",

year = "2017",

month = dec,

day = "4",

doi = "10.1109/PAC.2017.22",

language = "English",

series = "Proceedings - 2017 IEEE Symposium on Privacy-Aware Computing, PAC 2017",

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

pages = "192--193",

booktitle = "Proceedings - 2017 IEEE Symposium on Privacy-Aware Computing, PAC 2017",

}

Kim, J, Mo, YJ, Lee, W & Nyang, D 2017, Dynamic security-level maximization for stabilized parallel deep learning architectures in surveillance applications. in Proceedings - 2017 IEEE Symposium on Privacy-Aware Computing, PAC 2017. Proceedings - 2017 IEEE Symposium on Privacy-Aware Computing, PAC 2017, vol. 2017-January, Institute of Electrical and Electronics Engineers Inc., pp. 192-193, 1st IEEE Symposium on Privacy-Aware Computing, PAC 2017, Washington, United States, 1/08/17. https://doi.org/10.1109/PAC.2017.22

Dynamic security-level maximization for stabilized parallel deep learning architectures in surveillance applications. / Kim, Joongheon; Mo, Yeong Jong; Lee, Woojoo et al.
Proceedings - 2017 IEEE Symposium on Privacy-Aware Computing, PAC 2017. Institute of Electrical and Electronics Engineers Inc., 2017. p. 192-193 (Proceedings - 2017 IEEE Symposium on Privacy-Aware Computing, PAC 2017; Vol. 2017-January).

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

TY - GEN

T1 - Dynamic security-level maximization for stabilized parallel deep learning architectures in surveillance applications

AU - Kim, Joongheon

AU - Mo, Yeong Jong

AU - Lee, Woojoo

AU - Nyang, Daehun

PY - 2017/12/4

Y1 - 2017/12/4

N2 - This paper introduces a new surveillance platform which is equipped with multiple parallel deep learning frameworks. The deep learning frameworks are used for the face recognition of input image and video streams from CCTV cameras in security applications. Each deep learning framework has its own accuracy (related to recognition performance) and operation time (related to system stability) those are in tradeoff relationship. Based on this system architecture, a new dynamic control algorithm which selects one deep learning framework for time-average security-level (i.e., machine learning accuracy for recognition and classification) maximization under the consideration of system stability. The performance of the proposed algorithm was evaluated and also verified that it achieves desired performance.

AB - This paper introduces a new surveillance platform which is equipped with multiple parallel deep learning frameworks. The deep learning frameworks are used for the face recognition of input image and video streams from CCTV cameras in security applications. Each deep learning framework has its own accuracy (related to recognition performance) and operation time (related to system stability) those are in tradeoff relationship. Based on this system architecture, a new dynamic control algorithm which selects one deep learning framework for time-average security-level (i.e., machine learning accuracy for recognition and classification) maximization under the consideration of system stability. The performance of the proposed algorithm was evaluated and also verified that it achieves desired performance.

KW - Deep Learning

KW - Lyapunov Optimization

KW - Security

UR - http://www.scopus.com/inward/record.url?scp=85046550289&partnerID=8YFLogxK

U2 - 10.1109/PAC.2017.22

DO - 10.1109/PAC.2017.22

M3 - Conference contribution

AN - SCOPUS:85046550289

T3 - Proceedings - 2017 IEEE Symposium on Privacy-Aware Computing, PAC 2017

SP - 192

EP - 193

BT - Proceedings - 2017 IEEE Symposium on Privacy-Aware Computing, PAC 2017

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 1st IEEE Symposium on Privacy-Aware Computing, PAC 2017

Y2 - 1 August 2017 through 3 August 2017

ER -

Kim J, Mo YJ, Lee W, Nyang D. Dynamic security-level maximization for stabilized parallel deep learning architectures in surveillance applications. In Proceedings - 2017 IEEE Symposium on Privacy-Aware Computing, PAC 2017. Institute of Electrical and Electronics Engineers Inc. 2017. p. 192-193. (Proceedings - 2017 IEEE Symposium on Privacy-Aware Computing, PAC 2017). doi: 10.1109/PAC.2017.22

Dynamic security-level maximization for stabilized parallel deep learning architectures in surveillance applications

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Fingerprint

Cite this