Spatio-Temporal Split Learning

Joongheon Kim; Seunghoon Park; Soyi Jung; Seehwan Yoo

doi:10.1109/DSN-S52858.2021.00016

Spatio-Temporal Split Learning

Joongheon Kim
, Seunghoon Park
, Soyi Jung
, Seehwan Yoo

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

9 Citations (Scopus)

Abstract

This paper proposes a novel split learning framework with multiple end-systems in order to realize privacy-preserving deep neural network computation. In conventional split learning frameworks, deep neural network computation is separated into multiple computing systems for hiding entire network architectures. In our proposed framework, multiple computing end-systems are sharing one centralized server in split learning computation, where the multiple end-systems are with input and first hidden layers and the centralized server is with the other hidden layers and output layer. This framework, which is called as spatio-Temporal split learning, is spatially separated for gathering data from multiple end-systems and also temporally separated due to the nature of split learning. Our performance evaluation verifies that our proposed framework shows near-optimal accuracy while preserving data privacy.

Original language	English
Title of host publication	Proceedings - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	11-12
Number of pages	2
ISBN (Electronic)	9781665435666
DOIs	https://doi.org/10.1109/DSN-S52858.2021.00016
Publication status	Published - 2021 Jun
Event	51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021 - Virtual, Taipei, Taiwan, Province of China Duration: 2021 Jun 21 → 2021 Jun 24

Publication series

Name	Proceedings - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021

Conference

Conference	51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021
Country/Territory	Taiwan, Province of China
City	Virtual, Taipei
Period	21/6/21 → 21/6/24

Bibliographical note

Publisher Copyright:
© 2021 IEEE.

Keywords

deep learning
privacy-preserving
Split learning

ASJC Scopus subject areas

Artificial Intelligence
Computer Networks and Communications
Safety, Risk, Reliability and Quality
Control and Optimization

Access to Document

10.1109/DSN-S52858.2021.00016

Cite this

Kim, J., Park, S., Jung, S., & Yoo, S. (2021). Spatio-Temporal Split Learning. In Proceedings - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021 (pp. 11-12). (Proceedings - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/DSN-S52858.2021.00016

Spatio-Temporal Split Learning. / Kim, Joongheon; Park, Seunghoon; Jung, Soyi et al.
Proceedings - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021. Institute of Electrical and Electronics Engineers Inc., 2021. p. 11-12 (Proceedings - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021).

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

Kim, J, Park, S, Jung, S & Yoo, S 2021, Spatio-Temporal Split Learning. in Proceedings - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021. Proceedings - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021, Institute of Electrical and Electronics Engineers Inc., pp. 11-12, 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021, Virtual, Taipei, Taiwan, Province of China, 21/6/21. https://doi.org/10.1109/DSN-S52858.2021.00016

Kim J, Park S, Jung S, Yoo S. Spatio-Temporal Split Learning. In Proceedings - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021. Institute of Electrical and Electronics Engineers Inc. 2021. p. 11-12. (Proceedings - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021). doi: 10.1109/DSN-S52858.2021.00016

Kim, Joongheon ; Park, Seunghoon ; Jung, Soyi et al. / Spatio-Temporal Split Learning. Proceedings - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021. Institute of Electrical and Electronics Engineers Inc., 2021. pp. 11-12 (Proceedings - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021).

@inproceedings{1102eee926b1407e94acf53643401de5,

title = "Spatio-Temporal Split Learning",

abstract = "This paper proposes a novel split learning framework with multiple end-systems in order to realize privacy-preserving deep neural network computation. In conventional split learning frameworks, deep neural network computation is separated into multiple computing systems for hiding entire network architectures. In our proposed framework, multiple computing end-systems are sharing one centralized server in split learning computation, where the multiple end-systems are with input and first hidden layers and the centralized server is with the other hidden layers and output layer. This framework, which is called as spatio-Temporal split learning, is spatially separated for gathering data from multiple end-systems and also temporally separated due to the nature of split learning. Our performance evaluation verifies that our proposed framework shows near-optimal accuracy while preserving data privacy.",

keywords = "deep learning, privacy-preserving, Split learning",

author = "Joongheon Kim and Seunghoon Park and Soyi Jung and Seehwan Yoo",

note = "Publisher Copyright: {\textcopyright} 2021 IEEE.; 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021 ; Conference date: 21-06-2021 Through 24-06-2021",

year = "2021",

month = jun,

doi = "10.1109/DSN-S52858.2021.00016",

language = "English",

series = "Proceedings - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021",

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

pages = "11--12",

booktitle = "Proceedings - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021",

}

TY - GEN

T1 - Spatio-Temporal Split Learning

AU - Kim, Joongheon

AU - Park, Seunghoon

AU - Jung, Soyi

AU - Yoo, Seehwan

PY - 2021/6

Y1 - 2021/6

N2 - This paper proposes a novel split learning framework with multiple end-systems in order to realize privacy-preserving deep neural network computation. In conventional split learning frameworks, deep neural network computation is separated into multiple computing systems for hiding entire network architectures. In our proposed framework, multiple computing end-systems are sharing one centralized server in split learning computation, where the multiple end-systems are with input and first hidden layers and the centralized server is with the other hidden layers and output layer. This framework, which is called as spatio-Temporal split learning, is spatially separated for gathering data from multiple end-systems and also temporally separated due to the nature of split learning. Our performance evaluation verifies that our proposed framework shows near-optimal accuracy while preserving data privacy.

AB - This paper proposes a novel split learning framework with multiple end-systems in order to realize privacy-preserving deep neural network computation. In conventional split learning frameworks, deep neural network computation is separated into multiple computing systems for hiding entire network architectures. In our proposed framework, multiple computing end-systems are sharing one centralized server in split learning computation, where the multiple end-systems are with input and first hidden layers and the centralized server is with the other hidden layers and output layer. This framework, which is called as spatio-Temporal split learning, is spatially separated for gathering data from multiple end-systems and also temporally separated due to the nature of split learning. Our performance evaluation verifies that our proposed framework shows near-optimal accuracy while preserving data privacy.

KW - deep learning

KW - privacy-preserving

KW - Split learning

UR - https://www.scopus.com/pages/publications/85114150393

U2 - 10.1109/DSN-S52858.2021.00016

DO - 10.1109/DSN-S52858.2021.00016

M3 - Conference contribution

AN - SCOPUS:85114150393

T3 - Proceedings - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021

SP - 11

EP - 12

BT - Proceedings - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks - Supplemental Volume, DSN-S 2021

Y2 - 21 June 2021 through 24 June 2021

ER -

Spatio-Temporal Split Learning

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this