Painting Outside the Box: Image Outpainting with GANs

TL;DR

This paper introduces a deep learning method using GANs for image outpainting, enabling the extension of images beyond their original boundaries with promising results, and employs a three-phase training schedule for stability.

Contribution

It presents a novel deep learning approach for image outpainting using GANs with a three-phase training schedule and local discriminators, demonstrating feasible and promising results.

Findings

Able to outpaint 128x128 images realistically

Supports recursive outpainting for extended images

Deep learning approaches are effective for image outpainting

Abstract

The challenging task of image outpainting (extrapolation) has received comparatively little attention in relation to its cousin, image inpainting (completion). Accordingly, we present a deep learning approach based on Iizuka et al. for adversarially training a network to hallucinate past image boundaries. We use a three-phase training schedule to stably train a DCGAN architecture on a subset of the Places365 dataset. In line with Iizuka et al., we also use local discriminators to enhance the quality of our output. Once trained, our model is able to outpaint $128 \times 128$ color images relatively realistically, thus allowing for recursive outpainting. Our results show that deep learning approaches to image outpainting are both feasible and promising.