Deep Learning-based Image Super-Resolution Considering Quantitative and Perceptual Quality

TL;DR

This paper introduces a deep learning-based super-resolution method that balances quantitative accuracy and perceptual naturalness, improving image quality without sacrificing traditional metrics.

Contribution

It proposes a novel multi-pass upscaling framework with discriminator and score predictor networks to enhance perceptual quality while maintaining quantitative performance.

Findings

Achieves better balance between quantitative and perceptual quality.

Outperforms existing super-resolution methods in experiments.

Demonstrates improved naturalness of upscaled images.

Abstract

Recently, it has been shown that in super-resolution, there exists a tradeoff relationship between the quantitative and perceptual quality of super-resolved images, which correspond to the similarity to the ground-truth images and the naturalness, respectively. In this paper, we propose a novel super-resolution method that can improve the perceptual quality of the upscaled images while preserving the conventional quantitative performance. The proposed method employs a deep network for multi-pass upscaling in company with a discriminator network and two quantitative score predictor networks. Experimental results demonstrate that the proposed method achieves a good balance of the quantitative and perceptual quality, showing more satisfactory results than existing methods.