Learnable Encryption with a Diffusion Property

Ijaz Ahmad; Joongheon Kim; Seokjoo Shin

doi:10.1109/DSN-S65789.2025.00015

Learnable Encryption with a Diffusion Property

Ijaz Ahmad
, Joongheon Kim
, Seokjoo Shin^*

^*Corresponding author for this work

School of Electrical Engineering

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

Abstract

Guaranteeing privacy in outsourced deep learning (DL) model training ensures client control over their data and eases burden on service providers. Although learnable encryption can enable DL training in the encryption domain, its existing approaches do not adhere to cryptographic standards. Therefore, we propose a learnable encryption function that processes plaintext in a way that confusion and diffusion properties are ensured. Simulations on a COVID-19 dataset confirm its efficacy.

Original language	English
Title of host publication	Proceedings - 2025 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025
Editors	Marcello Cinque, Domenico Cotroneo, Luigi De Simone, Matthias Eckhart, Patrick P. C. Lee, Saman Zonouz
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	259-260
Number of pages	2
ISBN (Electronic)	9798331512033
DOIs	https://doi.org/10.1109/DSN-S65789.2025.00015
Publication status	Published - 2025
Event	55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025 - Naples, Italy Duration: 2025 Jun 23 → 2025 Jun 26

Publication series

Name	Proceedings - 2025 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025

Conference

Conference	55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025
Country/Territory	Italy
City	Naples
Period	25/6/23 → 25/6/26

Bibliographical note

Publisher Copyright:
© 2025 IEEE.

Keywords

deep learning
learnable encryption

ASJC Scopus subject areas

Software
Artificial Intelligence
Computer Networks and Communications
Information Systems
Safety, Risk, Reliability and Quality

Access to Document

10.1109/DSN-S65789.2025.00015

Cite this

Ahmad, I., Kim, J., & Shin, S. (2025). Learnable Encryption with a Diffusion Property. In M. Cinque, D. Cotroneo, L. De Simone, M. Eckhart, P. P. C. Lee, & S. Zonouz (Eds.), Proceedings - 2025 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025 (pp. 259-260). (Proceedings - 2025 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/DSN-S65789.2025.00015

Learnable Encryption with a Diffusion Property. / Ahmad, Ijaz; Kim, Joongheon; Shin, Seokjoo.
Proceedings - 2025 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025. ed. / Marcello Cinque; Domenico Cotroneo; Luigi De Simone; Matthias Eckhart; Patrick P. C. Lee; Saman Zonouz. Institute of Electrical and Electronics Engineers Inc., 2025. p. 259-260 (Proceedings - 2025 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025).

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

Ahmad, I, Kim, J & Shin, S 2025, Learnable Encryption with a Diffusion Property. in M Cinque, D Cotroneo, L De Simone, M Eckhart, PPC Lee & S Zonouz (eds), Proceedings - 2025 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025. Proceedings - 2025 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025, Institute of Electrical and Electronics Engineers Inc., pp. 259-260, 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025, Naples, Italy, 25/6/23. https://doi.org/10.1109/DSN-S65789.2025.00015

Ahmad I, Kim J, Shin S. Learnable Encryption with a Diffusion Property. In Cinque M, Cotroneo D, De Simone L, Eckhart M, Lee PPC, Zonouz S, editors, Proceedings - 2025 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025. Institute of Electrical and Electronics Engineers Inc. 2025. p. 259-260. (Proceedings - 2025 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025). doi: 10.1109/DSN-S65789.2025.00015

Ahmad, Ijaz ; Kim, Joongheon ; Shin, Seokjoo. / Learnable Encryption with a Diffusion Property. Proceedings - 2025 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025. editor / Marcello Cinque ; Domenico Cotroneo ; Luigi De Simone ; Matthias Eckhart ; Patrick P. C. Lee ; Saman Zonouz. Institute of Electrical and Electronics Engineers Inc., 2025. pp. 259-260 (Proceedings - 2025 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025).

@inproceedings{aa358cc5cd704ffaaa39ea565f39e0cd,

title = "Learnable Encryption with a Diffusion Property",

abstract = "Guaranteeing privacy in outsourced deep learning (DL) model training ensures client control over their data and eases burden on service providers. Although learnable encryption can enable DL training in the encryption domain, its existing approaches do not adhere to cryptographic standards. Therefore, we propose a learnable encryption function that processes plaintext in a way that confusion and diffusion properties are ensured. Simulations on a COVID-19 dataset confirm its efficacy.",

keywords = "deep learning, learnable encryption",

author = "Ijaz Ahmad and Joongheon Kim and Seokjoo Shin",

note = "Publisher Copyright: {\textcopyright} 2025 IEEE.; 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025 ; Conference date: 23-06-2025 Through 26-06-2025",

year = "2025",

doi = "10.1109/DSN-S65789.2025.00015",

language = "English",

series = "Proceedings - 2025 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025",

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

pages = "259--260",

editor = "Marcello Cinque and Domenico Cotroneo and \{De Simone\}, Luigi and Matthias Eckhart and Lee, \{Patrick P. C.\} and Saman Zonouz",

booktitle = "Proceedings - 2025 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025",

}

TY - GEN

T1 - Learnable Encryption with a Diffusion Property

AU - Ahmad, Ijaz

AU - Kim, Joongheon

AU - Shin, Seokjoo

PY - 2025

Y1 - 2025

N2 - Guaranteeing privacy in outsourced deep learning (DL) model training ensures client control over their data and eases burden on service providers. Although learnable encryption can enable DL training in the encryption domain, its existing approaches do not adhere to cryptographic standards. Therefore, we propose a learnable encryption function that processes plaintext in a way that confusion and diffusion properties are ensured. Simulations on a COVID-19 dataset confirm its efficacy.

AB - Guaranteeing privacy in outsourced deep learning (DL) model training ensures client control over their data and eases burden on service providers. Although learnable encryption can enable DL training in the encryption domain, its existing approaches do not adhere to cryptographic standards. Therefore, we propose a learnable encryption function that processes plaintext in a way that confusion and diffusion properties are ensured. Simulations on a COVID-19 dataset confirm its efficacy.

KW - deep learning

KW - learnable encryption

UR - https://www.scopus.com/pages/publications/105011416628

U2 - 10.1109/DSN-S65789.2025.00015

DO - 10.1109/DSN-S65789.2025.00015

M3 - Conference contribution

AN - SCOPUS:105011416628

T3 - Proceedings - 2025 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025

SP - 259

EP - 260

BT - Proceedings - 2025 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025

A2 - Cinque, Marcello

A2 - Cotroneo, Domenico

A2 - De Simone, Luigi

A2 - Eckhart, Matthias

A2 - Lee, Patrick P. C.

A2 - Zonouz, Saman

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 55th Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2025

Y2 - 23 June 2025 through 26 June 2025

ER -

Learnable Encryption with a Diffusion Property

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this