Arbitrary-Scale Image Synthesis

TL;DR

This paper introduces a new method for generating images at any scale, including unseen ones, by designing scale-invariant positional encodings and using inter-scale augmentations, resulting in flexible and high-quality image synthesis.

Contribution

The paper proposes scale-consistent positional encodings and inter-scale augmentations, enabling arbitrary-scale image synthesis with improved perceptual quality at unseen scales.

Findings

Effective arbitrary-scale image generation demonstrated on multiple datasets.

Outperforms previous methods limited to discrete scales.

Maintains perceptual quality across a continuum of scales.

Abstract

Positional encodings have enabled recent works to train a single adversarial network that can generate images of different scales. However, these approaches are either limited to a set of discrete scales or struggle to maintain good perceptual quality at the scales for which the model is not trained explicitly. We propose the design of scale-consistent positional encodings invariant to our generator's layers transformations. This enables the generation of arbitrary-scale images even at scales unseen during training. Moreover, we incorporate novel inter-scale augmentations into our pipeline and partial generation training to facilitate the synthesis of consistent images at arbitrary scales. Lastly, we show competitive results for a continuum of scales on various commonly used datasets for image synthesis.