Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure Synthetic Data

Wang, Xintao; Xie, Liangbin; Dong, Chao; Shan, Ying

doi:10.48550/arxiv.2107.10833

Cited by 23 publications

(31 citation statements)

References 55 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To test the performance of SwinIR for realworld SR, we re-train SwinIR by using the same degradation model as BSRGAN for low-quality image synthesis. Since there is no ground-truth high-quality images, we only provide visual comparison with representative bicubic model ESRGAN [81] and state-of-the-art realworld image SR models RealSR [37], BSRGAN [89] and Real-ESRGAN [80]. As shown in Fig.…”

Section: Results On Image Srmentioning

confidence: 99%

See 1 more Smart Citation

SwinIR: Image Restoration Using Swin Transformer

Liang¹,

Cao²,

Sun³

et al. 2021

Preprint

View full text Add to dashboard Cite

Image restoration is a long-standing low-level vision problem that aims to restore high-quality images from lowquality images (e.g., downscaled, noisy and compressed images). While state-of-the-art image restoration methods are based on convolutional neural networks, few attempts have been made with Transformers which show impressive performance on high-level vision tasks. In this paper, we propose a strong baseline model SwinIR for image restoration based on the Swin Transformer. SwinIR consists of three parts: shallow feature extraction, deep feature extraction and high-quality image reconstruction. In particular, the deep feature extraction module is composed of several residual Swin Transformer blocks (RSTB), each of which has several Swin Transformer layers together with a residual connection. We conduct experiments on three representative tasks: image super-resolution (including classical, lightweight and real-world image super-resolution), image denoising (including grayscale and color image denoising) and JPEG compression artifact reduction. Experimental results demonstrate that SwinIR outperforms state-of-the-art methods on different tasks by up to 0.14∼0.45dB, while the total number of parameters can be reduced by up to 67%.

show abstract

Section: Results On Image Srmentioning

confidence: 99%

“…For classical and lightweight image SR, we only use the naive L 1 pixel loss as same as previous work to show the effectiveness of the proposed network. For real-world image SR, we use a combination of pixel loss, GAN loss and perceptual loss [81,89,80,27,39,81] to improve visual quality.…”

Section: Methodsmentioning

confidence: 99%

SwinIR: Image Restoration Using Swin Transformer

Liang¹,

Cao²,

Sun³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…However, the reconstruction accuracy of the above methods greatly depends on the accuracy of the degradation mode estimation. To address this issue, more implicit degradation modeling methods are proposed [35], [141], [142], which aim to implicitly learn the potential degradation modes by the external datasets.…”

Section: Blind Sisrmentioning

confidence: 99%

From Beginner to Master: A Survey for Deep Learning-based Single-Image Super-Resolution

Li¹,

Pei²,

Zeng³

2021

Preprint

View full text Add to dashboard Cite

Single-image super-resolution (SISR) is an important task in image processing, which aims to enhance the resolution of imaging systems. Recently, SISR has made a huge leap and has achieved promising results with the help of deep learning (DL). In this survey, we give an overview of DL-based SISR methods and group them according to their targets, such as reconstruction efficiency, reconstruction accuracy, and perceptual accuracy. Specifically, we first introduce the problem definition, research background, and the significance of SISR. Secondly, we introduce some related works, including benchmark datasets, upsampling methods, optimization objectives, and image quality assessment methods. Thirdly, we provide a detailed investigation of SISR and give some domain-specific applications of it. Fourthly, we present the reconstruction results of some classic SISR methods to intuitively know their performance. Finally, we discuss some issues that still exist in SISR and summarize some new trends and future directions. This is an exhaustive survey of SISR, which can help researchers better understand SISR and inspire more exciting research in this field. An investigation project for SISR is provided in https://github.com/CV-JunchengLi/SISR-Survey.

show abstract

“…IKC [14] and DAN [25] further propose to jointly perform degradation prediction and restoration in an iterative manner. Recent works also attempt to address blind SR within a one-branch networks [40,47]. There are few related works investigating the relationship between two-branch and one-branch networks for blind SR.…”

Section: Related Workmentioning

confidence: 99%

Finding Discriminative Filters for Specific Degradations in Blind Super-Resolution

Xie

Wang

Dong

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Recent blind super-resolution (SR) methods typically consist of two branches, one for degradation prediction and the other for conditional restoration. However, our experiments show that a one-branch network can achieve comparable performance to the two-branch scheme. Then we wonder: how can one-branch networks automatically learn to distinguish degradations? To find the answer, we propose a new diagnostic tool -Filter Attribution method based on Integral Gradient (FAIG). Unlike previous integral gradient methods, our FAIG aims at finding the most discriminative filters instead of input pixels/features for degradation removal in blind SR networks. With the discovered filters, we further develop a simple yet effective method to predict the degradation of an input image. Based on FAIG, we show that, in one-branch blind SR networks, 1) we are able to find a very small number of (1%) discriminative filters for each specific degradation; 2) The weights, locations and connections of the discovered filters are all important to determine the specific network function. 3) The task of degradation prediction can be implicitly realized by these discriminative filters without explicit supervised learning. Our findings can not only help us better understand network behaviors inside one-branch blind SR networks, but also provide guidance on designing more efficient architectures and diagnosing networks for blind SR.

show abstract

Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure Synthetic Data

Cited by 23 publications

References 55 publications

SwinIR: Image Restoration Using Swin Transformer

SwinIR: Image Restoration Using Swin Transformer

From Beginner to Master: A Survey for Deep Learning-based Single-Image Super-Resolution

Finding Discriminative Filters for Specific Degradations in Blind Super-Resolution

Contact Info

Product

Resources

About