DRFN: Deep Recurrent Fusion Network for Single-Image Super-Resolution with Large Factors

TL;DR

This paper introduces DRFN, a deep recurrent fusion network that improves single-image super-resolution by replacing traditional upsampling with transposed convolution and integrating multi-level features for better accuracy and visual quality.

Contribution

The paper presents a novel deep recurrent fusion network that uses transposed convolution and multi-level feature fusion for enhanced large-factor super-resolution.

Findings

DRFN outperforms most current methods in accuracy and visual quality.

It is especially effective for large-scale images.

Uses fewer parameters than comparable models.

Abstract

Recently, single-image super-resolution has made great progress owing to the development of deep convolutional neural networks (CNNs). The vast majority of CNN-based models use a pre-defined upsampling operator, such as bicubic interpolation, to upscale input low-resolution images to the desired size and learn non-linear mapping between the interpolated image and ground truth high-resolution (HR) image. However, interpolation processing can lead to visual artifacts as details are over-smoothed, particularly when the super-resolution factor is high. In this paper, we propose a Deep Recurrent Fusion Network (DRFN), which utilizes transposed convolution instead of bicubic interpolation for upsampling and integrates different-level features extracted from recurrent residual blocks to reconstruct the final HR images. We adopt a deep recurrence learning strategy and thus have a larger receptive field, which is conducive to reconstructing an image more accurately. Furthermore, we show that the multi-level fusion structure is suitable for dealing with image super-resolution problems. Extensive benchmark evaluations demonstrate that the proposed DRFN performs better than most current deep learning methods in terms of accuracy and visual effects, especially for large-scale images, while using fewer parameters.