InfinityGAN: Towards Infinite-Pixel Image Synthesis

TL;DR

InfinityGAN introduces a scalable, patch-based framework for generating arbitrarily large, realistic images with consistent global and local features, enabling new applications like style fusion and outpainting.

Contribution

The paper presents InfinityGAN, a novel patch-wise image synthesis method that efficiently generates large, high-quality images with global and local consistency, surpassing previous limitations.

Findings

Generates arbitrarily large images with high realism.

Achieves global and local consistency in large images.

Enables applications like style fusion and outpainting.

Abstract

We present a novel framework, InfinityGAN, for arbitrary-sized image generation. The task is associated with several key challenges. First, scaling existing models to an arbitrarily large image size is resource-constrained, in terms of both computation and availability of large-field-of-view training data. InfinityGAN trains and infers in a seamless patch-by-patch manner with low computational resources. Second, large images should be locally and globally consistent, avoid repetitive patterns, and look realistic. To address these, InfinityGAN disentangles global appearances, local structures, and textures. With this formulation, we can generate images with spatial size and level of details not attainable before. Experimental evaluation validates that InfinityGAN generates images with superior realism compared to baselines and features parallelizable inference. Finally, we show several applications unlocked by our approach, such as spatial style fusion, multi-modal outpainting, and image inbetweening. All applications can be operated with arbitrary input and output sizes. Please find the full version of the paper at https://openreview.net/forum?id=ufGMqIM0a4b .