Diffusion-GAN: Training GANs with Diffusion

TL;DR

Diffusion-GAN introduces a diffusion process into GAN training, using a timestep-dependent discriminator and adaptive diffusion to improve stability and realism in generated images, outperforming existing GAN methods.

Contribution

The paper proposes a novel Diffusion-GAN framework that integrates a diffusion chain and timestep-dependent discriminator, providing theoretical guarantees and improved training stability.

Findings

Produces more realistic images than baseline GANs

Achieves higher training stability and data efficiency

Outperforms state-of-the-art GANs on various datasets

Abstract

Generative adversarial networks (GANs) are challenging to train stably, and a promising remedy of injecting instance noise into the discriminator input has not been very effective in practice. In this paper, we propose Diffusion-GAN, a novel GAN framework that leverages a forward diffusion chain to generate Gaussian-mixture distributed instance noise. Diffusion-GAN consists of three components, including an adaptive diffusion process, a diffusion timestep-dependent discriminator, and a generator. Both the observed and generated data are diffused by the same adaptive diffusion process. At each diffusion timestep, there is a different noise-to-data ratio and the timestep-dependent discriminator learns to distinguish the diffused real data from the diffused generated data. The generator learns from the discriminator's feedback by backpropagating through the forward diffusion chain, whose length is adaptively adjusted to balance the noise and data levels. We theoretically show that the discriminator's timestep-dependent strategy gives consistent and helpful guidance to the generator, enabling it to match the true data distribution. We demonstrate the advantages of Diffusion-GAN over strong GAN baselines on various datasets, showing that it can produce more realistic images with higher stability and data efficiency than state-of-the-art GANs.