Model Parameter Estimation and Inference on Encrypted Domain: Application to Noise Reduction in Encrypted Images

Saetbyeol Lee, Ji Won Yoon

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

One of the major issues in security is how to protect the privacy of multimedia big data on cloud systems. Homomorphic Encryption (HE) is increasingly regarded as a way to maintain user privacy on the untrusted cloud. However, HE is not widely used in machine learning and signal processing communities because the HE libraries are currently supporting only simple operations like integer addition and multiplication. It is known that division and other advanced operations cannot feasibly be designed and implemented in HE libraries. Therefore, we propose a novel approach to building a practical matrix inversion operation using approximation theory on HE. The approximated inversion operation is applied to reduce unwanted noise on encrypted images. Our research also suggests the efficient computation techniques for encrypted matrices. We conduct the experiment with real binary images using open source library of HE.

Original languageEnglish
Title of host publicationInformation Security Applications - 18th International Conference, WISA 2017, Revised Selected Papers
EditorsBrent ByungHoon Kang, Taesoo Kim
PublisherSpringer Verlag
Pages115-126
Number of pages12
ISBN (Print)9783319935621
DOIs
Publication statusPublished - 2018 Jan 1
Event18th World International Conference on Information Security and Application, WISA 2017 - Jeju Island, Korea, Republic of
Duration: 2017 Aug 242017 Aug 26

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10763 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other18th World International Conference on Information Security and Application, WISA 2017
CountryKorea, Republic of
CityJeju Island
Period17/8/2417/8/26

Fingerprint

Homomorphic Encryption
Noise Reduction
Noise abatement
Parameter estimation
Cryptography
Parameter Estimation
Privacy
Model
Approximation theory
Matrix Inversion
Binary images
Binary Image
Approximation Theory
Open Source
Multimedia
Learning systems
Signal Processing
Division
Inversion
Machine Learning

Keywords

  • Cloud security
  • Homomorphic encryption
  • Image processing
  • Leveled fully homomorphic encryption
  • Statistical analysis

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Lee, S., & Yoon, J. W. (2018). Model Parameter Estimation and Inference on Encrypted Domain: Application to Noise Reduction in Encrypted Images. In B. B. Kang, & T. Kim (Eds.), Information Security Applications - 18th International Conference, WISA 2017, Revised Selected Papers (pp. 115-126). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10763 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-93563-8_10

Model Parameter Estimation and Inference on Encrypted Domain : Application to Noise Reduction in Encrypted Images. / Lee, Saetbyeol; Yoon, Ji Won.

Information Security Applications - 18th International Conference, WISA 2017, Revised Selected Papers. ed. / Brent ByungHoon Kang; Taesoo Kim. Springer Verlag, 2018. p. 115-126 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10763 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Lee, S & Yoon, JW 2018, Model Parameter Estimation and Inference on Encrypted Domain: Application to Noise Reduction in Encrypted Images. in BB Kang & T Kim (eds), Information Security Applications - 18th International Conference, WISA 2017, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10763 LNCS, Springer Verlag, pp. 115-126, 18th World International Conference on Information Security and Application, WISA 2017, Jeju Island, Korea, Republic of, 17/8/24. https://doi.org/10.1007/978-3-319-93563-8_10
Lee S, Yoon JW. Model Parameter Estimation and Inference on Encrypted Domain: Application to Noise Reduction in Encrypted Images. In Kang BB, Kim T, editors, Information Security Applications - 18th International Conference, WISA 2017, Revised Selected Papers. Springer Verlag. 2018. p. 115-126. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-93563-8_10
Lee, Saetbyeol ; Yoon, Ji Won. / Model Parameter Estimation and Inference on Encrypted Domain : Application to Noise Reduction in Encrypted Images. Information Security Applications - 18th International Conference, WISA 2017, Revised Selected Papers. editor / Brent ByungHoon Kang ; Taesoo Kim. Springer Verlag, 2018. pp. 115-126 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{473f29c3c0c74495ac0c42329cc243a8,
title = "Model Parameter Estimation and Inference on Encrypted Domain: Application to Noise Reduction in Encrypted Images",
abstract = "One of the major issues in security is how to protect the privacy of multimedia big data on cloud systems. Homomorphic Encryption (HE) is increasingly regarded as a way to maintain user privacy on the untrusted cloud. However, HE is not widely used in machine learning and signal processing communities because the HE libraries are currently supporting only simple operations like integer addition and multiplication. It is known that division and other advanced operations cannot feasibly be designed and implemented in HE libraries. Therefore, we propose a novel approach to building a practical matrix inversion operation using approximation theory on HE. The approximated inversion operation is applied to reduce unwanted noise on encrypted images. Our research also suggests the efficient computation techniques for encrypted matrices. We conduct the experiment with real binary images using open source library of HE.",
keywords = "Cloud security, Homomorphic encryption, Image processing, Leveled fully homomorphic encryption, Statistical analysis",
author = "Saetbyeol Lee and Yoon, {Ji Won}",
year = "2018",
month = "1",
day = "1",
doi = "10.1007/978-3-319-93563-8_10",
language = "English",
isbn = "9783319935621",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "115--126",
editor = "Kang, {Brent ByungHoon} and Taesoo Kim",
booktitle = "Information Security Applications - 18th International Conference, WISA 2017, Revised Selected Papers",

}

TY - GEN

T1 - Model Parameter Estimation and Inference on Encrypted Domain

T2 - Application to Noise Reduction in Encrypted Images

AU - Lee, Saetbyeol

AU - Yoon, Ji Won

PY - 2018/1/1

Y1 - 2018/1/1

N2 - One of the major issues in security is how to protect the privacy of multimedia big data on cloud systems. Homomorphic Encryption (HE) is increasingly regarded as a way to maintain user privacy on the untrusted cloud. However, HE is not widely used in machine learning and signal processing communities because the HE libraries are currently supporting only simple operations like integer addition and multiplication. It is known that division and other advanced operations cannot feasibly be designed and implemented in HE libraries. Therefore, we propose a novel approach to building a practical matrix inversion operation using approximation theory on HE. The approximated inversion operation is applied to reduce unwanted noise on encrypted images. Our research also suggests the efficient computation techniques for encrypted matrices. We conduct the experiment with real binary images using open source library of HE.

AB - One of the major issues in security is how to protect the privacy of multimedia big data on cloud systems. Homomorphic Encryption (HE) is increasingly regarded as a way to maintain user privacy on the untrusted cloud. However, HE is not widely used in machine learning and signal processing communities because the HE libraries are currently supporting only simple operations like integer addition and multiplication. It is known that division and other advanced operations cannot feasibly be designed and implemented in HE libraries. Therefore, we propose a novel approach to building a practical matrix inversion operation using approximation theory on HE. The approximated inversion operation is applied to reduce unwanted noise on encrypted images. Our research also suggests the efficient computation techniques for encrypted matrices. We conduct the experiment with real binary images using open source library of HE.

KW - Cloud security

KW - Homomorphic encryption

KW - Image processing

KW - Leveled fully homomorphic encryption

KW - Statistical analysis

UR - http://www.scopus.com/inward/record.url?scp=85049487943&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85049487943&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-93563-8_10

DO - 10.1007/978-3-319-93563-8_10

M3 - Conference contribution

AN - SCOPUS:85049487943

SN - 9783319935621

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 115

EP - 126

BT - Information Security Applications - 18th International Conference, WISA 2017, Revised Selected Papers

A2 - Kang, Brent ByungHoon

A2 - Kim, Taesoo

PB - Springer Verlag

ER -