High-Resolution Be Aware! Improving the Self-Supervised Real-World Super-Resolution

TL;DR

This paper enhances self-supervised real-world super-resolution by introducing a degradation-aware controller and a feature-alignment regularizer, leading to more natural and perceptually superior high-resolution images.

Contribution

It proposes a novel degradation-aware controller and a feature-alignment regularizer to improve self-supervised super-resolution in real-world scenarios.

Findings

Achieves state-of-the-art perceptual quality in super-resolved images.

Effectively models real-world degradations for more natural results.

Finetunes existing SR models for specific real-world domains.

Abstract

Self-supervised learning is crucial for super-resolution because ground-truth images are usually unavailable for real-world settings. Existing methods derive self-supervision from low-resolution images by creating pseudo-pairs or by enforcing a low-resolution reconstruction objective. These methods struggle with insufficient modeling of real-world degradations and the lack of knowledge about high-resolution imagery, resulting in unnatural super-resolved results. This paper strengthens awareness of the high-resolution image to improve the self-supervised real-world super-resolution. We propose a controller to adjust the degradation modeling based on the quality of super-resolution results. We also introduce a novel feature-alignment regularizer that directly constrains the distribution of super-resolved images. Our method finetunes the off-the-shelf SR models for a target real-world domain. Experiments show that it produces natural super-resolved images with state-of-the-art perceptual performance.