Nested Scale Editing for Conditional Image Synthesis

TL;DR

This paper introduces a nested scale editing method for conditional image synthesis that enables precise control and diversity across scales, outperforming existing methods in tasks like outpainting, superresolution, and translation.

Contribution

It presents a novel nested scale disentanglement loss and diversification constraint for improved stratified navigation and control in latent space.

Findings

Outperforms state-of-the-art in image outpainting

Achieves higher accuracy in superresolution tasks

Provides better control in cross-domain translation

Abstract

We propose an image synthesis approach that provides stratified navigation in the latent code space. With a tiny amount of partial or very low-resolution image, our approach can consistently out-perform state-of-the-art counterparts in terms of generating the closest sampled image to the ground truth. We achieve this through scale-independent editing while expanding scale-specific diversity. Scale-independence is achieved with a nested scale disentanglement loss. Scale-specific diversity is created by incorporating a progressive diversification constraint. We introduce semantic persistency across the scales by sharing common latent codes. Together they provide better control of the image synthesis process. We evaluate the effectiveness of our proposed approach through various tasks, including image outpainting, image superresolution, and cross-domain image translation.