# BAYHENN: Combining Bayesian Deep Learning and Homomorphic Encryption for Secure DNN Inference

## TL;DR

BAYHENN introduces a practical privacy-preserving method for deep neural network inference by integrating homomorphic encryption with Bayesian neural networks, effectively protecting client and server data while achieving significant latency improvements.

## Contribution

It is the first to combine homomorphic encryption with Bayesian neural networks for secure DNN inference, enhancing privacy and efficiency.

## Key findings

- Achieves about 5x faster latency than GAZELLE.
- Successfully protects both client data and server model privacy.
- Demonstrates effectiveness on MNIST and clinical datasets.

## Abstract

Recently, deep learning as a service (DLaaS) has emerged as a promising way to facilitate the employment of deep neural networks (DNNs) for various purposes. However, using DLaaS also causes potential privacy leakage from both clients and cloud servers. This privacy issue has fueled the research interests on the privacy-preserving inference of DNN models in the cloud service. In this paper, we present a practical solution named BAYHENN for secure DNN inference. It can protect both the client's privacy and server's privacy at the same time. The key strategy of our solution is to combine homomorphic encryption and Bayesian neural networks. Specifically, we use homomorphic encryption to protect a client's raw data and use Bayesian neural networks to protect the DNN weights in a cloud server. To verify the effectiveness of our solution, we conduct experiments on MNIST and a real-life clinical dataset. Our solution achieves consistent latency decreases on both tasks. In particular, our method can outperform the best existing method (GAZELLE) by about 5x, in terms of end-to-end latency.