Low Latency Privacy Preserving Inference

TL;DR

This paper introduces two novel methods to significantly reduce latency in privacy-preserving neural network inference using homomorphic encryption, enabling wider networks and faster processing for sensitive data applications.

Contribution

The paper presents over 10x latency improvements and transfer learning techniques for private inference, addressing key limitations of existing homomorphic encryption approaches.

Findings

Achieved over 10x latency reduction in privacy-preserving inference

Enabled inference on wider neural networks with the same security level

Demonstrated practical private inference with ~0.16 seconds latency on vision tasks

Abstract

When applying machine learning to sensitive data, one has to find a balance between accuracy, information security, and computational-complexity. Recent studies combined Homomorphic Encryption with neural networks to make inferences while protecting against information leakage. However, these methods are limited by the width and depth of neural networks that can be used (and hence the accuracy) and exhibit high latency even for relatively simple networks. In this study we provide two solutions that address these limitations. In the first solution, we present more than $10\times$ improvement in latency and enable inference on wider networks compared to prior attempts with the same level of security. The improved performance is achieved by novel methods to represent the data during the computation. In the second solution, we apply the method of transfer learning to provide private inference services using deep networks with latency of $\sim0.16$ seconds. We demonstrate the efficacy of our methods on several computer vision tasks.