CollaFuse: Collaborative Diffusion Models

TL;DR

CollaFuse introduces a collaborative diffusion model approach inspired by split learning, reducing client computational load and enhancing privacy in distributed image synthesis tasks.

Contribution

The paper presents a novel distributed collaborative diffusion model framework that alleviates client computational burdens and improves privacy by outsourcing expensive processes.

Findings

Enhanced performance on CelebA, CIFAR-10, and Animals-with-Attributes2 datasets.

Reduces information disclosure by minimizing raw data sharing.

Demonstrates potential for edge computing applications.

Abstract

In the landscape of generative artificial intelligence, diffusion-based models have emerged as a promising method for generating synthetic images. However, the application of diffusion models poses numerous challenges, particularly concerning data availability, computational requirements, and privacy. Traditional approaches to address these shortcomings, like federated learning, often impose significant computational burdens on individual clients, especially those with constrained resources. In response to these challenges, we introduce the novel approach CollaFuse for distributed collaborative diffusion models inspired by split learning. Our approach facilitates collaborative training of diffusion models while alleviating client computational burdens during image synthesis. This reduced computational burden is achieved by retaining data and computationally inexpensive processes locally at each client while outsourcing the computationally expensive processes to shared, more efficient server resources. Through experiments on the common datasets CelebA, CIFAR-10, and Animals-with-Attributes2, our approach demonstrates enhanced performance while decreasing information disclosure as it reduces the necessity for sharing raw data. These capabilities hold significant potential across various application areas, including the design of edge computing solutions. Thus, our work advances distributed machine learning by contributing to the evolution of collaborative diffusion models.