DomainStudio: Fine-Tuning Diffusion Models for Domain-Driven Image Generation using Limited Data

TL;DR

DomainStudio enables high-quality, diverse image generation in limited-data scenarios by fine-tuning diffusion models to learn domain-specific features while preserving source diversity.

Contribution

This work introduces DomainStudio, a novel method for adapting pre-trained diffusion models to target domains with limited data, maintaining diversity and enhancing quality.

Findings

Achieves superior quality and diversity compared to GAN-based methods.

Effectively reduces overfitting in conditional diffusion models.

First to realize unconditional few-shot image generation with diffusion models.

Abstract

Denoising diffusion probabilistic models (DDPMs) have been proven capable of synthesizing high-quality images with remarkable diversity when trained on large amounts of data. Typical diffusion models and modern large-scale conditional generative models like text-to-image generative models are vulnerable to overfitting when fine-tuned on extremely limited data. Existing works have explored subject-driven generation using a reference set containing a few images. However, few prior works explore DDPM-based domain-driven generation, which aims to learn the common features of target domains while maintaining diversity. This paper proposes a novel DomainStudio approach to adapt DDPMs pre-trained on large-scale source datasets to target domains using limited data. It is designed to keep the diversity of subjects provided by source domains and get high-quality and diverse adapted samples in target domains. We propose to keep the relative distances between adapted samples to achieve considerable generation diversity. In addition, we further enhance the learning of high-frequency details for better generation quality. Our approach is compatible with both unconditional and conditional diffusion models. This work makes the first attempt to realize unconditional few-shot image generation with diffusion models, achieving better quality and greater diversity than current state-of-the-art GAN-based approaches. Moreover, this work also significantly relieves overfitting for conditional generation and realizes high-quality domain-driven generation, further expanding the applicable scenarios of modern large-scale text-to-image models.