Security for Distributed Deep Neural Networks: Towards Data Confidentiality & Intellectual Property Protection

Laurent Gomez, Marcus Wilhelm, José Márquez, Patrick Duverger

2019

Abstract

Current developments in Enterprise Systems observe a paradigm shift, moving the needle from the backend to the edge sectors of those; by distributing data, decentralizing applications and integrating novel components seamlessly to the central systems. Distributively deployed AI capabilities will thrust this transition. Several non-functional requirements arise along with these developments, security being at the center of the discussions. Bearing those requirements in mind, hereby we propose an approach to holistically protect distributed Deep Neural Network (DNN) based/enhanced software assets, i.e. confidentiality of their input & output data streams as well as safeguarding their Intellectual Property. Making use of Fully Homomorphic Encryption (FHE), our approach enables the protection of Distributed Neural Networks, while processing encrypted data. On that respect we evaluate the feasibility of this solution on a Convolutional Neuronal Network (CNN) for image classification deployed on distributed infrastructures.

Paper Citation

in Harvard Style

Gomez L., Wilhelm M., Márquez J. and Duverger P. (2019). Security for Distributed Deep Neural Networks: Towards Data Confidentiality & Intellectual Property Protection.In Proceedings of the 16th International Joint Conference on e-Business and Telecommunications - Volume 2: SECRYPT, ISBN 978-989-758-378-0, pages 439-447. DOI: 10.5220/0007922404390447

in Bibtex Style

@conference{secrypt19,
author={Laurent Gomez and Marcus Wilhelm and José Márquez and Patrick Duverger},
title={Security for Distributed Deep Neural Networks: Towards Data Confidentiality & Intellectual Property Protection},
booktitle={Proceedings of the 16th International Joint Conference on e-Business and Telecommunications - Volume 2: SECRYPT,},
year={2019},
pages={439-447},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0007922404390447},
isbn={978-989-758-378-0},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 16th International Joint Conference on e-Business and Telecommunications - Volume 2: SECRYPT,
TI - Security for Distributed Deep Neural Networks: Towards Data Confidentiality & Intellectual Property Protection
SN - 978-989-758-378-0
AU - Gomez L.
AU - Wilhelm M.
AU - Márquez J.
AU - Duverger P.
PY - 2019
SP - 439
EP - 447
DO - 10.5220/0007922404390447