Few-Shot Adaptation of Generative Adversarial Networks

TL;DR

This paper introduces FSGAN, a novel method for adapting pre-trained GANs to new domains using very few images, significantly improving visual quality in few-shot image synthesis tasks.

Contribution

The paper presents a simple, effective approach for few-shot GAN adaptation by manipulating singular values of weights, enabling high-quality synthesis with limited data.

Findings

FSGAN outperforms existing methods in visual quality for few-shot adaptation.

The method demonstrates effectiveness with as few as 5 images.

Standard quantitative metrics may not fully capture synthesis quality in few-shot scenarios.

Abstract

Generative Adversarial Networks (GANs) have shown remarkable performance in image synthesis tasks, but typically require a large number of training samples to achieve high-quality synthesis. This paper proposes a simple and effective method, Few-Shot GAN (FSGAN), for adapting GANs in few-shot settings (less than 100 images). FSGAN repurposes component analysis techniques and learns to adapt the singular values of the pre-trained weights while freezing the corresponding singular vectors. This provides a highly expressive parameter space for adaptation while constraining changes to the pretrained weights. We validate our method in a challenging few-shot setting of 5-100 images in the target domain. We show that our method has significant visual quality gains compared with existing GAN adaptation methods. We report qualitative and quantitative results showing the effectiveness of our method. We additionally highlight a problem for few-shot synthesis in the standard quantitative metric used by data-efficient image synthesis works. Code and additional results are available at http://e-271.github.io/few-shot-gan.