CipherDM: Secure Three-Party Inference for Diffusion Model Sampling

TL;DR

CipherDM introduces a secure MPC-based framework for diffusion model sampling, significantly enhancing privacy protection while reducing computation and communication costs in image generation tasks.

Contribution

It is the first universal framework applying MPC to diffusion models, addressing efficiency issues and enabling secure sampling across multiple DM tasks.

Findings

Improves running time by up to 2.328 times compared to direct implementation.

Reduces communication costs by up to 1.791 times.

Successfully applied to popular architectures like DDPM and DDIM.

Abstract

Diffusion Models (DMs) achieve state-of-the-art synthesis results in image generation and have been applied to various fields. However, DMs sometimes seriously violate user privacy during usage, making the protection of privacy an urgent issue. Using traditional privacy computing schemes like Secure Multi-Party Computation (MPC) directly in DMs faces significant computation and communication challenges. To address these issues, we propose CipherDM, the first novel, versatile and universal framework applying MPC technology to DMs for secure sampling, which can be widely implemented on multiple DM based tasks. We thoroughly analyze sampling latency breakdown, find time-consuming parts and design corresponding secure MPC protocols for computing nonlinear activations including SoftMax, SiLU and Mish. CipherDM is evaluated on popular architectures (DDPM, DDIM) using MNIST dataset and on SD deployed by diffusers. Compared to direct implementation on SPU, our approach improves running time by approximately 1.084\times \sim 2.328\times, and reduces communication costs by approximately 1.212\times \sim 1.791\times.