Download This PaperA Neural Network Application of Fully Homomorphic Encryption for Cloud Computing
5 Pages Posted: 3 Apr 2020
Date Written: April 8, 2020
Abstract
Cloud computing is a very convenient way of processing huge databases while sharing resources. Currently there are many cloud services available at comparatively low costs, which encourages researchers in the field of Machine and Deep Learning to use them. The problem with clouds is their inherent insecure nature. Hence sensitive sectors like health care and banking, are hesitant to use these services. In this paper we provide a proof of concept for applying “Fully Homomorphic Encryption” to process images on clouds for Neural Network application. We demonstrate the success of FHE on black-and-white, grayscale and color images, Gaussian filtering and on Convolutional Neural Network (CNN) for image classification, to some extent.
Keywords: Fully homomorphic encryption, cloud computing, image processing, CNN, BFV
Suggested Citation: Suggested Citation