Deep Unfolding Network for Image Super-Resolution

Zhang, Kai; Gool, Luc Van; Timofte, Radu

doi:10.1109/cvpr42600.2020.00328

Cited by 434 publications

(284 citation statements)

References 65 publications

Supporting

Mentioning

278

Contrasting

Unclassified

Order By: Relevance

“…and where s denotes distinct block processing operator with element-wise multiplication, i.e., applying element-wise multiplication to the s × s distinct blocks of F(k), ⇓ s denotes distinct block downsampler, i.e., averaging the s × s distinct blocks [80]. It is easy to verify that ( 15) is a special case of (14) with s = 1.…”

Section: Single Image Super-resolution (Sisr)mentioning

confidence: 99%

Plug-and-Play Image Restoration With Deep Denoiser Prior

Zhang¹,

Li²,

Zuo

et al. 2022

IEEE Trans. Pattern Anal. Mach. Intell.

Self Cite

414

302

View full text Add to dashboard Cite

Recent works on plug-and-play image restoration have shown that a denoiser can implicitly serve as the image prior for model-based methods to solve many inverse problems. Such a property induces considerable advantages for plug-and-play image restoration (e.g., integrating the flexibility of model-based method and effectiveness of learning-based methods) when the denoiser is discriminatively learned via deep convolutional neural network (CNN) with large modeling capacity. However, while deeper and larger CNN models are rapidly gaining popularity, existing plug-and-play image restoration hinders its performance due to the lack of suitable denoiser prior. In order to push the limits of plug-and-play image restoration, we set up a benchmark deep denoiser prior by training a highly flexible and effective CNN denoiser. We then plug the deep denoiser prior as a modular part into a half quadratic splitting based iterative algorithm to solve various image restoration problems. We, meanwhile, provide a thorough analysis of parameter setting, intermediate results and empirical convergence to better understand the working mechanism. Experimental results on three representative image restoration tasks, including deblurring, super-resolution and demosaicing, demonstrate that the proposed plug-and-play image restoration with deep denoiser prior not only significantly outperforms other state-of-the-art model-based methods but also achieves competitive or even superior performance against state-of-the-art learning-based methods. The source code is available at https://github.com/cszn/DPIR.

show abstract

Section: Single Image Super-resolution (Sisr)mentioning

confidence: 99%

Plug-and-Play Image Restoration With Deep Denoiser Prior

Zhang¹,

Li²,

Zuo

et al. 2022

IEEE Trans. Pattern Anal. Mach. Intell.

Self Cite

414

302

View full text Add to dashboard Cite

show abstract

“…In order to reflect the superiority of this algorithm in practical applications, this section selects the advanced methods of deep learning algorithms [ 33 ] and our super-resolution reconstruction method for comparison experiments. At this stage, deep learning-based super-resolution algorithms require a large number of high-definition images as training samples.…”

Section: Experiments and Results Analysismentioning

confidence: 99%

Blur Kernel Estimation and Non-Blind Super-Resolution for Power Equipment Infrared Images by Compressed Sensing and Adaptive Regularization

Zhao

Liu

Wang

2021

Sensors

View full text Add to dashboard Cite

Infrared sensing technology is more and more widely used in the construction of power Internet of Things. However, due to cost constraints, it is difficult to achieve the large-scale installation of high-precision infrared sensors. Therefore, we propose a blind super-resolution method for infrared images of power equipment to improve the imaging quality of low-cost infrared sensors. If the blur kernel estimation and non-blind super-resolution are performed at the same time, it is easy to produce sub-optimal results, so we chose to divide the blind super-resolution into two parts. First, we propose a blur kernel estimation method based on compressed sensing theory, which accurately estimates the blur kernel through low-resolution images. After estimating the blur kernel, we propose an adaptive regularization non-blind super-resolution method to achieve the high-quality reconstruction of high-resolution infrared images. According to the final experimental demonstration, the blind super-resolution method we proposed can effectively reconstruct low-resolution infrared images of power equipment. The reconstructed image has richer details and better visual effects, which can provide better conditions for the infrared diagnosis of the power system.

show abstract

“…Through preliminary experiments, we adopt three fully connected layers with ReLU as the first two activation functions and Softplus as the last to implement CM, similar to [23].…”

Section: Dynamic Gradient Descent Module (Dgdm)mentioning

confidence: 99%

ISTA-NET⁺⁺: Flexible Deep Unfolding Network for Compressive Sensing

You

Xie

Zhang

2021

2021 IEEE International Conference on Multimedia and Expo (ICME)

View full text Add to dashboard Cite

While deep neural networks have achieved impressive success in image compressive sensing (CS), most of them lack flexibility when dealing with multi-ratio tasks and multiscene images in practical applications. To tackle these challenges, we propose a novel end-to-end flexible ISTAunfolding deep network, dubbed ISTA-Net ++ , with superior performance and strong flexibility. Specifically, by developing a dynamic unfolding strategy, our model enjoys the adaptability of handling CS problems with different ratios, i.e., multi-ratio tasks, through a single model. A cross-block strategy is further utilized to reduce blocking artifacts and enhance the CS recovery quality. Furthermore, we adopt a balanced dataset for training, which brings more robustness when reconstructing images of multiple scenes. Extensive experiments on four datasets show that ISTA-Net ++ achieves state-of-the-art results in terms of both quantitative metrics and visual quality. Considering its flexibility, effectiveness and practicability, our model is expected to serve as a suitable baseline in future CS research.

show abstract

Deep Unfolding Network for Image Super-Resolution

Cited by 434 publications

References 65 publications

Plug-and-Play Image Restoration With Deep Denoiser Prior

Plug-and-Play Image Restoration With Deep Denoiser Prior

Blur Kernel Estimation and Non-Blind Super-Resolution for Power Equipment Infrared Images by Compressed Sensing and Adaptive Regularization

ISTA-NET⁺⁺: Flexible Deep Unfolding Network for Compressive Sensing

Contact Info

Product

Resources

About

Deep Unfolding Network for Image Super-Resolution

Cited by 434 publications

References 65 publications

Plug-and-Play Image Restoration With Deep Denoiser Prior

Plug-and-Play Image Restoration With Deep Denoiser Prior

Blur Kernel Estimation and Non-Blind Super-Resolution for Power Equipment Infrared Images by Compressed Sensing and Adaptive Regularization

ISTA-NET++: Flexible Deep Unfolding Network for Compressive Sensing

Contact Info

Product

Resources

About

ISTA-NET⁺⁺: Flexible Deep Unfolding Network for Compressive Sensing