Distributed Generative Adversarial Net

TL;DR

This paper introduces Distributed-GAN, a privacy-preserving multi-user GAN training framework that allows users to collaboratively generate diverse samples without sharing data, aiming to enhance AI service offerings.

Contribution

It proposes a novel distributed GAN framework enabling multi-user local training and diverse sample generation without data sharing, addressing privacy concerns.

Findings

Supports multi-user local training

Generates diverse samples

Enhances privacy in GAN training

Abstract

Recently the Generative Adversarial Network has become a hot topic. Considering the application of GAN in multi-user environment, we propose Distributed-GAN. It enables multiple users to train with their own data locally and generates more diverse samples. Users don't need to share data with each other to avoid the leakage of privacy. In recent years, commercial companies have launched cloud platforms based on artificial intelligence to provide model for users who lack computing power. We hope our work can inspire these companies to provide more powerful AI services.