Super-resolution Using Constrained Deep Texture Synthesis

TL;DR

This paper introduces a super-resolution method that leverages deep texture synthesis to generate high-frequency details, overcoming oversmoothing issues in traditional approaches and enhancing image quality.

Contribution

It demonstrates that deep learning-based texture synthesis can be effectively integrated into super-resolution to produce more detailed and realistic images.

Findings

Improved visual quality in super-resolved images.

Effective transfer of high-frequency details across diverse textures.

Outperforms traditional super-resolution methods in detail preservation.

Abstract

Hallucinating high frequency image details in single image super-resolution is a challenging task. Traditional super-resolution methods tend to produce oversmoothed output images due to the ambiguity in mapping between low and high resolution patches. We build on recent success in deep learning based texture synthesis and show that this rich feature space can facilitate successful transfer and synthesis of high frequency image details to improve the visual quality of super-resolution results on a wide variety of natural textures and images.