CVPR2020超分辨方向文章总结（中）

CVPR2020所有论文：https://openaccess.thecvf.com/CVPR2020

1.Closed-loop Matters: Dual Regression Networks for Single Image Super-Resolution

problem

Two underlying limitations to existing SR methods:

1. Learning the mapping function from LR to HR images is typically an ill-posed problem, because there exist infinite HR images that can be downsampled to the same LR image. As a result, the space of the possible functions can be extremely large, which makes it hard to find a good solution.
2. Paired LR-HR data may be unavailable in real-world applications and the underlying degradation method is often unknown. For such a more general case, existing SR models often incur the adaptation problem and yield poor performance.

作者分析研究传统方法可能产生的两个局限性，提出DRN方法。

针对1：
How to reduce the possible space of the mapping functions to improve the training of SR models becomes an important problem.
针对2：
How to effectively exploit the unpaired data to adapt SR models to real-world applications becomes an urgent and important problem.

solution

Propose a dual regression scheme by introducing an additional constraint on LR data to reduce the space of the possible functions.

Specifically, besides the mapping from LR to HR images, we learn an additional dual regression mapping estimates the down-sampling kernel and reconstruct LR images, which forms a closed-loop to provide additional supervision.

More critically, since the dual regression process does not depend on HR images, we can directly learn from LR images. In this sense, we can easily adapt SR models to real-world data.

2.Deep Unfolding Network for Image Super-Resolution

problem

Different from model-based methods that can handle the SISR problem with different scale factors, blur kernels and noise levels under a unified MAP (maximum a posteriori) framework, learning-based methods generally lack such flexibility.

solution

Propose a deep unfolding super-resolution network (USRNet) to bridge the gap between learning-based methods and model-based methods.

1. Similar to model-based methods, USRNet can effectively handle the classical degradation model with different blur kernels, scale factors and noise levels via a single model.
2. Similar to learning-based methods, USRNet can be trained in an end-to-end fashion to guarantee effectiveness and efficiency.

3.Unpaired Image Super-Resolution using Pseudo-Supervision

problem

In most studies on learning-based image super-resolution (SR), the paired training dataset is created by downscaling high-resolution (HR) images with a predetermined operation (e.g., bicubic). However, these methods fail to super-resolve real-world low-resolution (LR) images, for which the degradation process is much more complicated and unknown.

sotulion

Propose an unpaired SR method using a generative adversarial network that does not require a paired/aligned training dataset.

Our network consists of an unpaired kernel/noise correction network and a pseudo-paired SR network.

The correction network removes noise and adjusts the kernel of the inputted LR image; then, the corrected clean LR image is upscaled by the SR network.
In the training phase, the correction network also produces a pseudo-clean LR image from the inputted HR image, and then a mapping from the pseudo-clean LR image to the inputted HR image is learned by the SR network in a paired manner.

网络组成：
1.纠错网络
2.假的配对SR网络

解决问题领域：unpaired SR problem

4.Structure-Preserving Super Resolution with Gradient Guidance

problem

Recent studies benefiting from generative adversarial network (GAN) have promoted the development ofSISR by recovering photo-realistic images. However, there are always undesired structural distortions in the recovered images.

solution

Propose a structure-preserving super resolution method to alleviate the above issue while maintaining the merits of GAN-based methods to generate perceptual-pleasant details.

Specifically, we exploit gradient maps of images to guide the recovery in two aspects.

On the one hand, we restore high-resolution gradient maps by a gradient branch to provide additional structure priors for the SR process.

On the other hand, we propose a gradient loss which imposes a second-order restriction on the super-resolved images.

Along with the previous image-space loss functions, the gradient-space objectives help generative networks concentrate more on geometric structures.

Moreover, our method is model-agnostic, which can be potentially used for off-the-shelf SR networks.

First to explicitly consider preserving geometric structures in GAN-based SR methods.

5.Image Super-Resolution with Cross-Scale Non-Local Attention and Exhaustive Self-Exemplars Mining

problem

Deep convolution-based single image super-resolution(SISR) networks embrace the benefits of learning from large-scale external image resources for local recovery, yet most existing works have ignored the long-range feature-wise similarities in natural images.
Some recent works have successfully leveraged this intrinsic feature correlation by exploring non-local attention modules.
However, none of the current deep models have studied another inherent property ofimages: cross-scale feature correlation.

solution

Propose the first Cross-Scale Non-Local (CS-NL) attention module with integration into a recurrent neural network.

By combining the new CS-NL prior with local and in-scale non-local priors in a powerful recurrent fusion cell, we can find more cross-scale feature correlations within a single low-resolution (LR) image.