Precise Approximation of Convolutional Neural Networks for Homomorphically Encrypted Data
Open Access
- 19 June 2023
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 62062-62076
Abstract
Homomorphic encryption (HE) is one of the representative solutions to privacy-preserving machine learning (PPML) classification enabling the server to classify private data of clients while guaranteeing privacy. This work focuses on PPML using word-wise fully homomorphic encryption (FHE). In order to implement deep learning on word-wise HE, the ReLU and max-pooling functions should be approximated by polynomials for homomorphic operations. Most of the previous studies focus on HE-friendly networks, which approximate the ReLU and max-pooling functions using low-degree polynomials. However, this approximation cannot support deeper neural networks due to large approximation errors in general and can classify only relatively small datasets. Thus, we propose a precise polynomial approximation technique, a composition of minimax approximate polynomials of low degrees for the ReLU and max-pooling functions. If we replace the ReLU and max-pooling functions with the proposed approximate polynomials, standard deep learning models such as ResNet and VGGNet can still be used without further modification for PPML on FHE. Even pre-trained parameters can be used without retraining, which makes the proposed method more practical. We approximate the ReLU and max-pooling functions in the ResNet-152 using the composition of minimax approximate polynomials of degrees 15, 27, and 29. Then, we succeed in classifying the plaintext ImageNet dataset with 77.52% accuracy, which is very close to the original model accuracy of 78.31%. Also, we obtain an accuracy of 87.90% for classifying the encrypted CIFAR-10 dataset in the ResNet-20 without any additional training.Keywords
All Related Versions
Funding Information
- National Research Foundation of Korea
- Korea government (NRF-2021R1A2C2011082)
- Ministry of Science and ICT, South Korea (RS-2023-00212103)
This publication has 21 references indexed in Scilit:
- CHET: an optimizing compiler for fully-homomorphic neural-network inferencing2019
- nGraph-HE2019
- TFHE: Fast Fully Homomorphic Encryption Over the TorusJournal of Cryptology, 2019
- Fast Homomorphic Evaluation of Deep Discretized Neural Networks2018
- Homomorphic Encryption for Arithmetic of Approximate Numbers2017
- Deep Residual Learning for Image Recognition2016
- Rethinking the Inception Architecture for Computer Vision2016
- Going deeper with convolutions2015
- ImageNet: A large-scale hierarchical image database2009
- On the Number of Nonscalar Multiplications Necessary to Evaluate PolynomialsSIAM Journal on Computing, 1973