深度学习图像超分 物理模拟图像超分

… has been introduced into the field of image super-resolution (… current image SR algorithms based on deep learning. Firstly, … This paper focuses on single image super-resolution (SISR), …

ABSTRACT Remote sensing image super-resolution has become a critical task to enhance spatial details for downstream applications such as land cover mapping, environmental monitoring, and precision agriculture. However, the unique characteristics of remote sensing data – including limited training samples, complex sensor degradations, and spectral diversity – pose significant challenges to conventional super-resolution pipelines. In this paper, we present a comprehensive review of recent advances in remote sensing single image super resolution (RSSISR), spanning across supervised, self-supervised, unsupervised, training-free, generative adversarial network-based, diffusion-based, and physically/statistically guided frameworks. We introduce a structured taxonomy that organizes these approaches and analyzes their strengths, limitations, and application domains. In addition, we identify key challenges in the field, including (1) data scarcity, (2) generalization gaps, (3) high computational cost, (4) lack of task-oriented evaluation, and (5) limited physical integration in degradation process modelling. By pinpointing these challenges, we outline promising research directions including (1) data augmentation and enhancement, (2) generative priors and large foundation models, (3) learning paradigms less reliant on supervision, (4) application-aware frameworks, (5) physics-guided modelling, and (6) multi-task learning. Case studies across real-world tasks such as crop monitoring, spatio-temporal image upscaling, flood mapping, and object detection further illustrate the practical utility of different RSSISR strategies. This review not only summarizes the current landscape but also provides a forward-looking perspective on the future of RSSISR.

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 design targets. 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 at https://github.com/CV-JunchengLi/SISR-Survey.

… This paper aims to provide a complete overview of current improvements in Single Image Super-Resolution (SISR) in the medical field through the use of DL techniques. We can …

Single-image super-resolution (SISR) is an important task in image processing, aiming to achieve enhanced image resolution. With the development of deep learning, SISR based on convolutional neural networks has also gained great progress, but as the network deepens and the task of SISR becomes more complex, SISR networks become difficult to train, which hinders SISR from achieving greater success. Therefore, to further promote SISR, many challenges have emerged in recent years. In this review, we briefly review the SISR challenges organized from 2017 to 2022 and focus on the in-depth classification of these challenges, the datasets employed, the evaluation methods used, and the powerful network architectures proposed or accepted by the winners. First, depending on the tasks of the challenges, the SISR challenges can be broadly classified into four categories: classic SISR, efficient SISR, perceptual extreme SISR, and real-world SISR. Second, we introduce the datasets commonly used in the challenges in recent years and describe their characteristics. Third, we present the image evaluation methods commonly used in SISR challenges in recent years. Fourth, we introduce the network architectures used by the winners, mainly to explore in depth where the advantages of their network architectures lie and to compare the results of previous years’ winners. Finally, we summarize the methods that have been widely used in SISR in recent years and suggest several possible promising directions for future SISR.

… recent findings of single image super-resolution using deep learning with an emphasis … image super-resolution., it is also to highlight the potential applications of image super-resolution …

… encompassing single image super resolution (SISR) and multiple image super resolution (MISR… In SISR methods, addressing individual images independently, we review blind and non-…

Single Image Super Resolution (SSIR) is a problem in computer vision where the goal is 1 to create high-resolution images from low-resolution ones. It has important applications in fields 2 such as medical imaging and security surveillance. While traditional methods such as interpolation 3 and reconstruction-based models have been used in the past, deep learning techniques have recently 4 gained attention due to their superior performance and computational efficiency. This article proposes 5 an Autoencoder based Deep Learning Model for SSIR, in particular, a light model that uses fewer 6 parameters without compromising performance. The down-sampling part of the Autoencoder 7 mainly uses 3 by 3 convolution and has no subsampling layers. The up-sampling part uses transpose 8 convolution and residual connections from the down sampling part. The model is trained using a 9 subset of the VILRC ImageNet database. The model is evaluated using quantitative metrics PSNR, 10 SSIM as well as qualitative measures such as perceptual quality. PSNR and SSIM figures as high as 11 76.06 and 0.93 are reported.

… Single Image Super- Resolution (SISR) is a complex restoration method to recover high-resolution (HR) image from … The advantages of CNNs over other deep learning methods are: …

… Single Image Super-Resolution (SISR) is a fundamental computer vision task aimed at enhancing the spatial resolution and quality of low-resolution images… , a regression deep learning …

… Due to the boom of deep neural networks, several CNN- and transformer-based SR models [11, … Enhanced deep residual networks for single image super-resolution. In CVPRW, pages …

Single image super resolution has significantly advanced by utilizing transformers-based deep learning algorithms. However, challenges still need to be addressed in handling grid-like image patches with higher computational demands and addressing issues like over-smoothing in visual patches. This paper presents a Deep Learning model for single-image super-resolution. In this paper, we present the XTNSR model, a novel multi-path network architecture that combines Local feature window transformers (LWFT) with Xception blocks for single-image super-resolution. The model processes grid-like image patches effectively and reduces computational complexity by integrating a Patch Embedding layer. Whereas the Xception blocks use depth-wise separable convolutions for hierarchical feature extraction, the LWFT blocks capture long-range dependencies and fine-grained qualities. A multi-layer feature fusion block with skip connections, part of this hybrid architecture, guarantees efficient local and global feature fusion. The experimental results show better performance in Peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), and visual quality than the state-of-the-art techniques. By optimizing parameters, the suggested architecture also lowers computational complexity. Overall, the architecture presents a promising approach for advancing image super-resolution capabilities.

The current advancements in image super-resolution have explored different attention mechanisms to achieve better quantitative and perceptual results. The critical challenge recently is to utilize the potential of attention mechanisms to reconstruct high-resolution images from their low-resolution counterparts. This research proposes a novel method that combines inception blocks, non-local sparse attention, and a U-Net network architecture. The network incorporates the non-local sparse attention on the backbone of symmetric encoder-decoder U-Net structure, which helps to identify long-range dependencies and exploits contextual information while preserving global context. By incorporating skip connections, the network can leverage features at different scales, enhancing the reconstruction of high-frequency information. Additionally, we introduce inception blocks allowing the model to capture information at various levels of abstraction to enhance multi-scale representation learning further. Experimental findings show that our suggested approach produces superior quantitative measurements, such as peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), visual information fidelity (VIF), and visually appealing high-resolution image reconstructions.

… Single Image Super-Resolution (SISR) has made significant advancements with both CNN-… improvement of 0.28 dB for × 2 lightweight super-resolution on Urban100, with a reduction in …

Single-molecule localization microscopy (SMLM) can be used to resolve subcellular structures and achieve a tenfold improvement in spatial resolution compared to that obtained by conventional fluorescence microscopy. However, the separation of single-molecule fluorescence events that requires thousands of frames dramatically increases the image acquisition time and phototoxicity, impeding the observation of instantaneous intracellular dynamics. Here we develop a deep-learning based single-frame super-resolution microscopy (SFSRM) method which utilizes a subpixel edge map and a multicomponent optimization strategy to guide the neural network to reconstruct a super-resolution image from a single frame of a diffraction-limited image. Under a tolerable signal density and an affordable signal-to-noise ratio, SFSRM enables high-fidelity live-cell imaging with spatiotemporal resolutions of 30 nm and 10 ms, allowing for prolonged monitoring of subcellular dynamics such as interplays between mitochondria and endoplasmic reticulum, the vesicle transport along microtubules, and the endosome fusion and fission. Moreover, its adaptability to different microscopes and spectra makes it a useful tool for various imaging systems.

… It is anticipated that the SISR machine learning method did not receive enough results for the image super-resolution (SR) [1]. Because deep learning has succeeded in other computer …

… Image SuperResolution (SISR), which is a type of super-resolution … The SISR method works by utilizing information from low… Image Super-Resolution (CNN-SISR). This approach is a …

… Neural Networks (CNN) are … Super-Resolution (SISR) task. However, the optimal integration of these two components to fully exploit their complementary strengths for improving SISR …

The primary goal of Single Image Super-Resolution (SISR), a fundamental yet challenging computer vision task with several practical applications in domains such as surveillance, medical imaging, and remote sensing, is to reconstruct a high-resolution (HR) image from a single low- resolution (LR) input. The performance of SISR has been greatly improved by the advent of deep learning, specifically Convolutional Neural Networks (CNNs) and Transformer architectures. An extensive review of deep learning-based SISR techniques is presented in this study. Begin by formulating the SISR problem and discussing prevalent evaluation metrics that balance distortion (e.g., PSNR) and perceptual quality (e.g., SSIM, LPIPS). Subsequently, classifying and analyzing key methodologies across five categories: interpolation-based and traditional models, CNN-based architectures (e.g., SRCNN, VDSR, EDSR), GAN-based frameworks (e.g., SRGAN, ESRGAN, Real-ESRGAN), attention-enhanced networks (e.g., RCAN), and Transformer-based approaches (e.g., SwinIR, HAT). In each category, the theoretical framework, design innovations, and corresponding advantages and limitations are explored. By showing architectural design strategies and training paradigms, this review highlights a structured understanding of the significant evolution from early CNNs to sophisticated GANs and Transformers in SISR, serving as a reference for future model development and practical deployment.

Deep 4 learning (DL)-based single image super-resolution (SISR) for low-resolution (LR) images typically aims to recover a high-resolution (HR) image from its LR version due to downsampling and blurring imperfections of the imaging sensor. The existing DL SR networks reasonably solve the downsampling problem, however, they do not address the complex deblurring problem, simultaneously. To address the later, we propose a joint dual-branch convolutional neural network (CNN) for recovering sharp HR images from LR images degraded with Gaussian blur. The proposed method has two task-independent networks: 1) super-resolution (SR) and 2) deblurring. In particular, we adopt a residual spatial and channel squeeze-and-excitation (RSCSE) module incorporating concurrent spatial and channel squeeze-and-excitation (SCSE) attention mechanism and local feature fusion (LFF) concepts in the SR network. Furthermore, the deblurring network is designed based on a SCSE-based encoder-decoder module to retrieve sharp features from blurred LR images. The feature maps obtained from these networks are adaptively fused by learning a gated module with attention mechanism to generate a clear HR. Experimental results demonstrate that the proposed method outperforms other state-of-the-art DL techniques in visual results and quantitative metrics; peak signalto-noise ratio (PSNR) improves by 1.4 dB-4.9 dB and 0.4 dB-2.6 dB for zooming factors 2 and 4, respectively, on publicly available RGB remote sensing (RS) datasets. Similarly, for multispectral (MS) datasets, they are 1.4 dB-3.5 dB and 0.2 dB-1.4 dB for zooming factors 2 and 4, respectively. It also provides promising results for land cover classification in RS applications.

Super resolution (SR) has garnered significant attention within the computer vision community, driven by advances in deep learning (DL) techniques and the growing demand for high-quality visual applications. With the expansion of this field, numerous surveys have emerged. Most existing surveys focus on specific domains, lacking a comprehensive overview of this field. Here, we present an in-depth review of diverse SR methods, encompassing single-image SR (SISR), video SR (VSR), stereo SR (SSR), and light field SR (LFSR). We extensively cover over 150 SISR methods, nearly 70 VSR approaches, and approximately 30 techniques for SSR and LFSR. We analyze methodologies, datasets, evaluation protocols, empirical results, and complexity. In addition, we conducted a taxonomy based on each backbone structure according to the diverse purposes. We also explore valuable yet understudied open issues in the field. We believe that this work will serve as a valuable resource and offer guidance to researchers in this domain. To facilitate access to related work, we created a dedicated repository available at https://github.com/AVC2-UESTC/Holistic-Super-Resolution-Review

The subject of this article is Image Super-Resolution (ISR) using deep learning techniques.  ISR is a rapidly evolving research area in computer science that focuses on producing high-resolution images from one or more low-resolution sources. It has garnered substantial interest due to its broad applications in areas such as medical imaging, remote sensing, and multimedia. The rise of deep learning techniques has brought a revolution in ISR, providing superior performance and computational efficiency compared to traditional methods and driving further advancements in overcoming the challenges associated with enhancing image resolution. The goal of this study is to enhance the quality of super-resolved images by developing a novel deep learning approach. Specifically, we explore the integration of Convolutional Neural Networks (CNNs) and Generative Adversarial Networks (GANs) to address the inherent challenges of producing high-quality images from low-resolution data. This study aims to push the boundaries of ISR by combining these architectures for greater precision and visual fidelity. The tasks are as follows: 1) design and implement a hybrid model using CNNs and GANs for image super-resolution tasks; 2) train the model on benchmark datasets like Set5, Set14, DIV2K, and specialized datasets such as X-ray images; 3) assess the model’s performance using numerical metrics like Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM); 4) to compare the proposed method against existing state-of-the-art ISR techniques and demonstrate its superiority. The following results were obtained in this study: Our deep learning model, which integrates the Super-Resolution Convolutional Neural Network (SRCNN) and the Super-Resolution Generative Adversarial Network (SRGAN), demonstrated significant performance improvements. The CNN successfully learned to map low-resolution image patches to their high-resolution counterparts, and the GAN further refined the images, enhancing both precision and visual quality. The evaluation metrics yielded highly promising results, with Peak Signal-to-Noise Ratio (PSNR) reaching up to 36.1368 dB and Structural Similarity Index Measure (SSIM) reaching 0.9670. These values exceed the benchmarks set by contemporary ISR methods, thus validating the superiority and effectiveness of our approach in the field of image super-resolution. Conclusions. This study demonstrated the potential of combining CNN and GAN in the domain of image super-resolution. The proposed model exhibits significant advancements over existing ISR methods, offering higher accuracy and improved image quality. The findings confirm the efficiency of deep learning methods in overcoming traditional imaging challenges, making the proposed model valuable for both academic research and practical applications in ISR.

Adapting the Diffusion Probabilistic Model (DPM) for direct image super-resolution is wasteful, given that a simple Convolutional Neural Network (CNN) can recover the main low-frequency content. Therefore, we present ResDiff, a novel Diffusion Probabilistic Model based on Residual structure for Single Image Super-Resolution (SISR). ResDiff utilizes a combination of a CNN, which restores primary low-frequency components, and a DPM, which predicts the residual between the ground-truth image and the CNN predicted image. In contrast to the common diffusion-based methods that directly use LR space to guide the noise towards HR space, ResDiff utilizes the CNN’s initial prediction to direct the noise towards the residual space between HR space and CNN-predicted space, which not only accelerates the generation process but also acquires superior sample quality. Additionally, a frequency-domain-based loss function for CNN is introduced to facilitate its restoration, and a frequency-domain guided diffusion is designed for DPM on behalf of predicting high-frequency details. The extensive experiments on multiple benchmark datasets demonstrate that ResDiff outperforms previous diffusion based methods in terms of shorter model convergence time, superior generation quality, and more diverse samples.

Single image super‐resolution (SISR) is a promising research direction in computer vision and image processing for improving the visual perception of low‐quality images. In recent years, deep learning algorithms have driven tremendous development in SR, and SR methods based on various network architectures have significantly improved the quality of reconstructed images. Although there has been a large amount of reviews focusing on SISR, few studies have focused specifically on network architectures for SISR. This paper aims to provide a systematic overview of the design ideas of SISR using multiple architectures, including Convolutional Neural Networks (CNN), Generative Adversarial Networks (GAN), Transformer, and Diffusion model. In addition, an experimental analysis and comparison of state‐of‐the‐art SR algorithms have been performed on publicly available quantitative and qualitative datasets. Finally, some future directions are discussed that may help other community researchers.

Recently, CNN-based methods for hyperspectral image super-resolution (HSISR) have achieved outstanding performance. Due to the multi-band property of hyperspectral images, 3D convolutions are natural candidates for extracting spatial–spectral correlations. However, pure 3D CNN models are rare to see, since they are generally considered to be too complex, require large amounts of data to train, and run the risk of overfitting on relatively small-scale hyperspectral datasets. In this paper, we question this common notion and propose Full 3D U-Net (F3DUN), a full 3D CNN model combined with the U-Net architecture. By introducing skip connections, the model becomes deeper and utilizes multi-scale features. Extensive experiments show that F3DUN can achieve state-of-the-art performance on HSISR tasks, indicating the effectiveness of the full 3D CNN on HSISR tasks, thanks to the carefully designed architecture. To further explore the properties of the full 3D CNN model, we develop a 3D/2D mixed model, a popular kind of model prior, called Mixed U-Net (MUN) which shares a similar architecture with F3DUN. Through analysis on F3DUN and MUN, we find that 3D convolutions give the model a larger capacity; that is, the full 3D CNN model can obtain better results than the 3D/2D mixed model with the same number of parameters when it is sufficiently trained. Moreover, experimental results show that the full 3D CNN model could achieve competitive results with the 3D/2D mixed model on a small-scale dataset, suggesting that 3D CNN is less sensitive to data scaling than what people used to believe. Extensive experiments on two benchmark datasets, CAVE and Harvard, demonstrate that our proposed F3DUN exceeds state-of-the-art HSISR methods both quantitatively and qualitatively.

Neural Radiance Fields (NeRF) have achieved great success in the task of synthesizing novel views that preserve the same resolution as the training views. However, it is challenging for NeRF to synthesize high-quality high-resolution novel views with low-resolution training data. To solve this problem, we propose a zero-shot super-resolution training framework for NeRF. This framework aims to guide the NeRF model to synthesize high-resolution novel views via single-scene internal learning rather than requiring any external high-resolution training data. Our approach consists of two stages. First, we learn a scene-specific degradation mapping by performing internal learning on a pretrained low-resolution coarse NeRF. Second, we optimize a super-resolution fine NeRF by conducting inverse rendering with our mapping function so as to backpropagate the gradients from low-resolution 2D space into the super-resolution 3D sampling space. Then, we further introduce a temporal ensemble strategy in the inference phase to compensate for the scene estimation errors. Our method is featured on two points: (1) it does not consume high-resolution views or additional scene data to train super-resolution NeRF; (2) it can speed up the training process by adopting a coarse-to-fine strategy. By conducting extensive experiments on public datasets, we have qualitatively and quantitatively demonstrated the effectiveness of our method.

… Utilizing image-specific internal priors for zero-shot learning … images, and existing zero-shot SISR methods that rely solely … issues of zero-shot methods, this paper proposes a Zero-Shot …

The asymmetric dual-lens configuration is commonly available on mobile devices nowadays, which naturally stores a pair of wide-angle and telephoto images of the same scene to support realistic super-resolution (SR). Even on the same device, however, the degradation for modeling realistic SR is image-specific due to the unknown acquisition process (e.g., tiny camera motion). In this paper, we propose a zero-shot solution for dual-lens SR (ZeDuSR), where only the dual-lens pair at test time is used to learn an image-specific SR model. As such, ZeDuSR adapts itself to the current scene without using external training data, and thus gets rid of generalization difficulty. However, there are two major challenges to achieving this goal: 1) dual-lens alignment while keeping the realistic degradation, and 2) effective usage of highly limited training data. To overcome these two challenges, we propose a degradation-invariant alignment method and a degradation-aware training strategy to fully exploit the information within a single dual-lens pair. Extensive experiments validate the superiority of Ze-DuSR over existing solutions on both synthesized and real-world dual-lens datasets. The implementation code is available at https://github.com/XrKang/ZeDuSR.

… subject-specific super-resolution without relying on … , Zero-Shot Super Resolution (ZSSR), on retrospective high-field and prospective low-field scans. The evaluation of the zero-shot SR …

Most Single Image Super-Resolution (SISR) meth-ods rely on paired training data, typically generated through bicubic downsampling, which fails to capture the complex degra-dations seen in real-world scenarios. This reliance on synthetic data creates a performance gap when deploying SISR models on actual degraded images. To address this limitation, we pro-pose a Robust Zero-Shot learning-based Generative Adversarial Network for Super-Resolution (RZSGAN-SR) framework that leverages zero-shot learning to handle unknown degradations effectively. Our approach incorporates a Zero-Shot learning-based Degradation Correction Network (ZSDCN) to translate real-world degraded Low-Resolution (LR) images into synthetic LR images with known degradations. These translated images are then fed into a lightweight, robust Generative Adversarial Network (GAN)-based SR network to generate high-quality, visually realistic Super-Resolved (SR) images. More specifically, the proposed RZSGA-SR is a two-phase framework consisting of zero-shot degradation correction and efficient GAN-based upsampling. This hybrid model leverages the adaptability of Zero-Shot Learning (ZSL) with the realism of a robust GAN-based SR network with high fidelity and perceptual quality SR reconstruction. Extensive experiments show that RZSGAN-SR surpasses state-of-the-art methods, achieving superior reconstruction (PSNR, SSIM) and perceptual quality (LPIPS) on real-world degraded images.

Zero-shot super-resolution (ZSSR) has generated a lot of interest due to its flexibility in various applications. However, the computational demands of ZSSR make it ineffective when dealing with large-scale low-resolution image sets. To address this issue, we propose a novel meta-learning model. We treat the set of low-resolution images as a collection of ZSSR tasks and learn meta-knowledge about ZSSR by leveraging these tasks. This approach reduces the computational burden of super-resolution for large-scale low-resolution images. Additionally, through multiple ZSSR task learning, we uncover a general super-resolution model that enhances the generalization capacity of ZSSR. Finally, using the learned meta-knowledge, our model achieves impressive results with just a few gradient updates when given a novel task. We evaluate our method using two remote sensing datasets with varying spatial resolutions. Our experimental results demonstrate that using multiple ZSSR tasks yields better outcomes than a single task, and our method outperforms other state-of-the-art super-resolution methods.

… Existing polarized image super-resolution (SR) … , a zero-shot learning network is proposed for polarized SR with little sample, called ZPSR (Zero-shot Polarization Super-Resolution). …

… , employing a Zero-Shot Super-Resolution Generative Adversarial … combines the strengths of zero-shot super-resolution (ZSSR) and … The zero-shot super-resolution model serves as the …

When the unknown degradation is mixed with unknown blurry kernels, how to perform super-resolution operation is an open issue. The mean idea of the existing zero-shot and non-zero-shot methods is to estimate blurry kernel. The effects of these methods depend on the accuracy of the deduced blurry kernel. In this paper, we propose Randomly initialized Zero-Shot Super-Resolution (RZSR) training strategy. RZSR is a zero-shot training method and it allows the network to extract low-resolution image features and generate its counterpart high-resolution images under the interference of degradation algorithms. We further propose two model-agnostic modules which are Adaptive Information Extraction Module (AIEM) and knowledge dictionary. They respectively assist the network to extract features and well fit the data distribution of clear images. RZSR can be applied to any single image super-resolution and video super-resolution models. We prove the generalization ability and superiority of RZSR through a series of experiments.

… In order to compare the performance of ZSSR in handling complex degraded images, we … ZSSR and ZSSR+KernelGAN. It can be seen that the PSNR and SSIM values of ZSSR alone …

… super-resolution (SR) methods typically rely on vast quantities of paired data. As an essential solution, zero-shot … In this work, we propose a novel guided Diffusion model for Zero-shot …

The advent of Deep Learning (DL) techniques has significantly improved the performance of Image Super-Resolution (ISR) algorithms. However, the primary limitation to extending the existing DL-based works for real- world instances is their computational and time complexities. Besides this, the presumed degradation process in their training datasets is another. In this paper, we present a lightweight and highly efficient zero-shot ISR model. The proposed algorithmfirst estimates the degradation kernel K from the given low-resolution (LR) image statistics. Later, we introduce "Deep Identity Learning (DIL)", a novel learning strategy, to compute the inverse of K by exploiting the identity relation between the degradation and inverse degradation models. Contrary to the mainstream ISR works, the proposed model considers K alone as its input to learn the ISR task. We term the proposed approach as "Image Specific Super-Resolution Using Deep Identity Learning (ISSR-DIL)". In our experiments, ISSR-DIL demonstrated a competitive performance compared to state-of- the-art (SotA) works on benchmark ISR datasets while requiring, at least by order of 10, fewer computational resources.

… Zero-Shot Super-Resolution (ZSSR) utilizes internal image statistics to improve resolution using only single images [8]. Other unsupervised techniques leverage self-supervision or …

Deep convolutional neural networks (DCNNs) have manifested significant performance gains for single-image super-resolution (SISR) in the past few years. Most of the existing methods are generally implemented in a fully supervised way using large-scale training samples and only learn the SR models restricted to specific data. Thus, the adaptation of these models to real low-resolution (LR) images captured under uncontrolled imaging conditions usually leads to poor SR results. This study proposes a zero-shot blind SR framework via leveraging the power of deep learning, but without the requirement of the prior training using predefined imaged samples. It is well known that there are two unknown data: the underlying target high-resolution (HR) images and the degradation operations in the imaging procedure hidden in the observed LR images. Taking these in mind, we specifically employed two deep networks for respectively modeling the priors of both the target HR image and its corresponding degradation kernel and designed a degradation block to realize the observation procedure of the LR image. Via formulating the loss function as the approximation error of the observed LR image, we established a completely blind end-to-end zero-shot learning framework for simultaneously predicting the target HR image and the degradation kernel without any external data. In particular, we adopted a multi-scale encoder–decoder subnet to serve as the image prior learning network, a simple fully connected subnet to serve as the kernel prior learning network, and a specific depthwise convolutional block to implement the degradation procedure. We conducted extensive experiments on several benchmark datasets and manifested the great superiority and high generalization of our method over both SOTA supervised and unsupervised SR methods.

… To address the above problems, an end-to-end fundus image SRR model called LiteZSSR is proposed, which improves upon the CNN structure of zero-shot super-resolution (ZSSR) [5] …

By automatically learning the priors embedded in images with powerful modelling capabilities, deep learning‐based algorithms have recently made considerable progress in reconstructing the high‐resolution hyperspectral (HR‐HS) image. With previously collected large‐amount of external data, these methods are intuitively realised under the full supervision of the ground‐truth data. Thus, the database construction in merging the low‐resolution (LR) HS (LR‐HS) and HR multispectral (MS) or RGB image research paradigm, commonly named as HSI SR, requires collecting corresponding training triplets: HR‐MS (RGB), LR‐HS and HR‐HS image simultaneously, and often faces difficulties in reality. The learned models with the training datasets collected simultaneously under controlled conditions may significantly degrade the HSI super‐resolved performance to the real images captured under diverse environments. To handle the above‐mentioned limitations, the authors propose to leverage the deep internal and self‐supervised learning to solve the HSI SR problem. The authors advocate that it is possible to train a specific CNN model at test time, called as deep internal learning (DIL), by on‐line preparing the training triplet samples from the observed LR‐HS/HR‐MS (or RGB) images and the down‐sampled LR‐HS version. However, the number of the training triplets extracted solely from the transformed data of the observation itself is extremely few particularly for the HSI SR tasks with large spatial upscale factors, which would result in limited reconstruction performance. To solve this problem, the authors further exploit deep self‐supervised learning (DSL) by considering the observations as the unlabelled training samples. Specifically, the degradation modules inside the network were elaborated to realise the spatial and spectral down‐sampling procedures for transforming the generated HR‐HS estimation to the high‐resolution RGB/LR‐HS approximation, and then the reconstruction errors of the observations were formulated for measuring the network modelling performance. By consolidating the DIL and DSL into a unified deep framework, the authors construct a more robust HSI SR method without any prior training and have great potential of flexible adaptation to different settings per observation. To verify the effectiveness of the proposed approach, extensive experiments have been conducted on two benchmark HS datasets, including the CAVE and Harvard datasets, and demonstrate the great performance gain of the proposed method over the state‐of‐the‐art methods.

… In this study, the ZSSR algorithm was used for super-resolution of images at scales from 2 up to 9, based on learning the internal statistics of an image as mentioned in [11]. The ZSSR …

As a medical imaging modality, many researches have been devoted to improving the resolution of optical coherence tomography (OCT). We developed a deep-learning based OCT self super-resolution (OCT-SSR) pipeline to improve the axial resolution of OCT images based on the high-resolution and low-resolution spectral data collected by the OCT system. In this pipeline, the enhanced super-resolution asymmetric generative adversarial networks were built to improve the network outputs without increasing the complexity. The feasibility and effectiveness of the approach were demonstrated by experimental results on the images of the biological samples collected by the home-made spectral-domain OCT and swept-source OCT systems. More importantly, we found the sidelobes in the original images can be obviously suppressed while improving the resolution based on the OCT-SSR method, which can help to reduce pseudo-signal in OCT imaging when non-Gaussian spectra light source is used. We believe that the OCT-SSR method has broad prospects in breaking the limitation of the source bandwidth on the axial resolution of the OCT system.

Although existing image deep learning super-resolution (SR) methods achieve promising performance on benchmark datasets, they still suffer from severe performance drops when the degradation of the low-resolution (LR) input is not covered in training. To address the problem, we propose an innovative unsupervised method of Learning Correction Filter via Degradation-Adaptive Regression for Blind Single Image Super-Resolution. Highly inspired by the generalized sampling theory, our method aims to enhance the strength of off-the-shelf SR methods trained on known degradations and adapt to unknown complex degradations to generate improved results. Specifically, we first conduct degradation estimation for each local image region by learning the internal distribution in an unsupervised manner via GAN. Instead of assuming degradation are spatially invariant across the whole image, we learn correction filters to adjust degradations to known degradations in a spatially variant way by a novel linearly-assembled pixel degradation-adaptive regression module (DARM). DARM is lightweight and easy to optimize on a dictionary of multiple pre-defined filter bases. Extensive experiments on synthetic and real-world datasets verify the effectiveness of our method both qualitatively and quantitatively. Code can be available at: https://github.com/edbca/DARSR.

In recent years, super-resolution reconstruction has been introduced into DEM. The process of mapping low-resolution DEM images to high-resolution DEM is highly uncertain. At present, DEM super-resolution reconstruction methods mainly solve the problem by designing a more sophisticated network. However, the existing methods fail to capture the complex conditional distribution of high-resolution DEM during training, resulting in blurring and artifacts in the reconstruction results. Based on the lack of explicit, high-resolution DEM conditional distribution modeling, this paper proposes a reversible network model based on normalized flow. The model uses the characteristics of real low-resolution DEM images as conditions and learns to map the distribution of high-resolution DEM images to simple Gaussian distribution, thereby simulating the conditional distribution of high-resolution DEM. The negative log-likelihood function and pixel loss function are used to accelerate the optimization to generate high-resolution DEM images that are closer to the natural terrain. Experiments show that the proposed model can preserve the terrain features and achieve good performance. Especially on the test set, compared with the traditional interpolation method (Bicubic) and the existing deep learning methods (SRGAN and Internal–External), the PSNR results of this model are improved by 2.03%, 0.43%, and 2.58%, respectively.

The adoption of Stereo Imaging technology within endoscopic procedures represents a transformative advancement in medical imaging, providing surgeons with depth perception and detailed views of internal anatomy for enhanced diagnostic accuracy and surgical precision. However, the practical application of stereo imaging in endoscopy faces challenges, including the generation of low-resolution and blurred images, which can hinder the effectiveness of medical diagnoses and interventions. Our research introduces an endoscopic image SR model in response to these specific. This model features an innovative feature extraction module and an advanced cross-view feature interaction module tailored for the intricacies of endoscopic imagery. Initially trained on the SCARED dataset, our model was rigorously tested across four additional publicly available endoscopic image datasets at scales 2, 4, and 8, demonstrating unparalleled performance improvements in endoscopic SR. Our results are compelling. They show that our model not only substantially enhances the quality of endoscopic images but also consistently surpasses other existing methods like E-SEVSR, DCSSRNet, and CCSBESR in all tested datasets, in quantitative measures such as PSNR and SSIM, and in qualitative evaluations. The successful application of our SR model in endoscopic imaging has the potential to revolutionize medical diagnostics and surgery, significantly increasing the precision and effectiveness of endoscopic procedures. The code will be released on GitHub and can be accessed at https://github.com/cu-vtrg-lab/Saliency-Aware-Deep-Learning-Approach-for-Enhanced-Endoscopic-Image-SR.

Despite efforts to construct super-resolution (SR) training datasets with a wide range of degradation scenarios, existing supervised methods based on these datasets still struggle to consistently offer promising results due to the diversity of real-world degradation scenarios and the inherent complexity of model learning. Our work explores a new route: integrating the sample-adaptive property learned through image intrinsic self-similarity and the universal knowledge acquired from large-scale data. We achieve this by uniting internal learning and external learning by an unrolled optimization process. With the merits of both, the tuned fully-supervised SR models can be augmented to broadly handle the real-world degradation in a plug-and-play style. Furthermore, to promote the efficiency of combining internal/external learning, we apply an attention-based weight-updating method to guide the mining of self-similarity, and various data augmentations are adopted while applying the exponential moving average strategy. We conduct extensive experiments on real-world degraded images and our approach outperforms other methods in both qualitative and quantitative comparisons. Our project is available at: https://github.com/ZahraFan/AdaSSR/.

Real-world degradations deviate from ideal degradations, as most deep learning-based scenarios involve the ideal synthesis of low-resolution (LR) counterpart images by popularly used bicubic interpolation. Moreover, supervised learning approaches rely on many high-resolution (HR) and LR image pairings to reconstruct missing information based on their association, developed by complex long hours of deep neural network training. Additionally, the trained model's generalizability on various image datasets with various distributions is not guaranteed. To overcome this challenge, we proposed our novel Self-FuseNet, particularly for extremely poor-resolution satellite images. Also, the network exhibits strong generalization performance on additional datasets (both “ideal” and “nonideal” scenarios). The network is especially for those image datasets suffering from the following two significant limitations: 1) nonavailability of ground truth HR images; 2) limitation of a large count of the unpaired dataset for deep neural network training. The benefit of the proposed model is threefold: 1) it does not require any significant extensive training data, either paired or unpaired but only a single LR image without prior knowledge of its distribution; 2) it is a simple and effective model for super-resolving very poor-resolution images, saving computational resources and time; 3) using UNet, the processing of data are accelerated by the network's wide skip connections, allowing image reconstruction with fewer parameters. Rather than using an inverse approach, as common in most deep learning scenarios, we introduced a forward approach to super-resolve exceptionally LR remote sensing images. This demonstrates its supremacy over recently proposed state-of-the-art methods for unsupervised single real-world image blind super-resolution.

Single image super-solution (SR) aims to restore a high-resolution (HR) image from a degraded low-resolution (LR) image. However, existing SR models still face a significant domain gap between synthetic and real-world datasets due to the mismatched degradation distributions, hindering SR models from achieving optimal results. In this paper, we propose an unsupervised diffusion-based degradation modeling framework (UDDM) to effectively capture real-world degradation distributions. Specifically, given unpaired LR and HR images, a diffusion-based degradation module (DDM) first models the degradation distribution by diffusing real-world LR images to downsampled LR images, which does not require HR images. It then applies reverse diffusion to generate real-world LR images from extremely downsampled HR images. This approach allows DDM to model and generate real-world degradation distributions without requiring paired data, by using extreme downsampling to link unpaired LR and HR images. Additionally, we introduce a physics-based dynamic degradation module (P-DDM) that adaptively models content-aware degradation, ensuring both content and structural accuracy. Finally, the LR images generated by DDM and P-DDM are adaptively weighted to produce the final LR images, which are paired with the given HR images for training the SR network. Extensive experiments across multiple real-world datasets demonstrate that our framework achieves state-of-the-art performance in both qualitative and quantitative comparison.

Inindustrial vision systems, image degradation due to noise, illumination variation, and optical aberrations undermines the reliability of downstream tasks, such as defect detection and parameter measurement. This article proposes PhySISR, a physics-consistent self-supervised single-image super-resolution framework tailored for industrial applications. It integrates region-aware noise simulation, Airy-disk point spread function-based degradation modeling, and structure-aware loss design to restore fine details and edge structures without any labeled data. A lightweight network is further designed to reduce parameters and inference latency, ensuring deployment efficiency. To validate the method, we construct and release SMT-ImageSet, a real-world industrial dataset captured from surface-mount equipment under diverse imaging conditions. Experiments demonstrate that PhySISR outperforms representative supervised and self-supervised methods in structural recovery, edge clarity, and downstream tasks, such as binarization and parameter extraction, showing strong applicability for practical industrial image enhancement.

… model for light propagation into a super-resolution DM. These … tures in the cell walls, owing to its embedded physics-based … -guided diffusion model for reversing the degradation in …

Image distortion by atmospheric turbulence is a stochastic degradation, which is a critical problem in long-range optical imaging systems. A number of research has been conducted during the past decades, including model-based and emerging deep-learning solutions with the help of synthetic data. Although fast and physics-grounded simulation tools have been introduced to help the deep-learning models adapt to real-world turbulence conditions recently, the training of such models only relies on the synthetic data and ground truth pairs. This paper proposes the Physics-integrated Restoration Network (PiRN) to bring the physics-based simulator directly into the training process to help the network to disentangle the stochasticity from the degradation and the underlying image. Furthermore, to overcome the "average effect" introduced by deterministic models and the domain gap between the synthetic and real-world degradation, we further introduce PiRN with Stochastic Refinement (PiRN-SR) to boost its perceptual quality. Overall, our PiRN and PiRN-SR improve the generalization to real-world unknown turbulence conditions and provide a state-of-the-art restoration in both pixel-wise accuracy and perceptual quality. Our codes are available at https://github.com/VITA-Group/PiRN.

Machine Learning, particularly Generative Adversarial Networks (GANs), has revolutionised Super-Resolution (SR). However, generated images often lack physical meaningfulness, which is essential for scientific applications. Our approach, PC-SRGAN, enhances image resolution while ensuring physical consistency for interpretable simulations. PC-SRGAN significantly improves both the Peak Signal-to-Noise Ratio and the Structural Similarity Index Measure compared to conventional SR methods, even with limited training data (e.g., only 13% of training data is required to achieve performance similar to SRGAN). Beyond SR, PC-SRGAN augments physically meaningful machine learning, incorporating numerically justified time integrators and advanced quality metrics. These advancements promise reliable and causal machine-learning models in scientific domains. A significant advantage of PC-SRGAN over conventional SR techniques is its physical consistency, which makes it a viable surrogate model for time-dependent problems. PC-SRGAN advances scientific machine learning by improving accuracy and efficiency, enhancing process understanding, and broadening applications to scientific research.

Fourier single-pixel detection (FSPD) suffers from an inherent tradeoff between spatial resolution and detection efficiency. This constraint significantly impacts measurement precision in optical instrumentation applications. To overcome these measurement limitations, we propose a frequency-domain super-resolution network based on physical measurement principles tailored for FSPD instruments. This method decomposes the frequency domain into magnitude and phase branches with targeted optimization strategies based on Fourier imaging distribution characteristics while leveraging the physical sparse measurement priors of FSPD frequency sampling to enhance detection accuracy and measurement noise suppression. The network utilizes residual dense blocks (RDBs) for low-frequency measurement stability and combines self-attention (SA) mechanisms with nonlocal (NL) blocks for high-frequency detail detection. By integrating measurement consistency constraints and physics-informed frequency weighting based on FSPD’s inherent sampling characteristics, the proposed approach achieves superior measurement reconstruction accuracy. The physical measurement priors in the frequency domain are used to constrain the structural consistency of the super-resolution detection process. A multiloss function with physical measurement constraints balances spectral fidelity and measurement consistency. Experiments demonstrate that under $128\times 128$ resolution undersampling detection, we achieved $256\times 256$ resolution high-accuracy measurements. Moreover, at the same resolution, high-precision detection has been achieved for challenging measurement scenarios, such as A4 paper reflective media, which are difficult to reconstruct with traditional FSPD instruments, advancing practical measurement applications, including precision microscopy detection and scattering measurement systems.

Integrating deep learning with fluid dynamics presents a promising path for advancing the comprehension of complex flow phenomena within both theoretical and practical engineering domains. Despite this potential, considerable challenges persist, particularly regarding the calibration and training of deep learning models. This paper conducts an extensive review and analysis of recent developments in deep learning architectures that aim to enhance the accuracy of fluid flow data interpretation. It investigates various applications, architectural designs, and performance evaluation metrics. The analysis covers several models, including convolutional neural networks, generative adversarial networks, physics-informed neural networks, transformer models, diffusion models, and reinforcement learning frameworks, emphasizing components improving reconstruction capabilities. Standard performance metrics are employed to rigorously evaluate the models' reliability and efficacy in producing high-performance results applicable across spatiotemporal flow data. The findings emphasize the essential role of deep learning in representing fluid flows and address ongoing challenges related to the systems' high degrees of freedom, precision demands, and resilience to error.

Deep learning superresolution (SR) is a promising approach to reduce MRI scan time without requiring custom sequences or iterative reconstruction. Previous deep learning SR approaches have generated low‐resolution training images by simple k‐space truncation, but this does not properly model in‐plane turbo spin echo (TSE) MRI resolution degradation, which has variable T2 relaxation effects in different k‐space regions. To fill this gap, we developed a T2‐deblurred deep learning SR method for the SR of 3D‐TSE images.

Abstract Deep learning-based image super resolution (SR) is an image processing technique designed to enhance the resolution of digital images. With the continuous improvement of methods and the growing availability of large real-world datasets, this technology has gained significant importance in a wide variety of research fields in recent years. In this paper, we present a comprehensive review of promising developments in deep learning-based image super resolution. First, we give an overview of contributions outside the field of microscopy before focusing on the specific application areas of light optical microscopy, fluorescence microscopy and scanning electron microscopy. Using selected examples, we demonstrate how the application of deep learning-based image super resolution techniques has resulted in substantial improvements to specific use cases. Additionally, we provide a structured analysis of the architectures used, evaluation metrics, error functions, and more. Finally, we discuss current trends, existing challenges, and offer guidance for selecting suitable methods.

… an urban flood event in Portland, Oregon (physics-based simulation scenario). FLO-SR … model generalization. Despite these limitations, FLO-SR combined with physics-based modeling …

Recent image degradation estimation methods have enabled single-image super-resolution (SR) approaches to better upsample real-world images. Among these methods, explicit kernel estimation approaches have demonstrated unprecedented performance at handling unknown degradations. Nonetheless, a number of limitations constrain their efficacy when used by downstream SR models. Specifically, this family of methods yields i) excessive inference time due to long per-image adaptation times and ii) inferior image fidelity due to kernel mismatch. In this work, we introduce a learning-to-learn approach that meta-learns from the information contained in a distribution of images, thereby enabling significantly faster adaptation to new images with substantially improved performance in both kernel estimation and image fidelity. Specifically, we meta-train a kernelgenerating GAN, named MetaKernelGAN, on a range of tasks, such that when a new image is presented, the generator starts from an informed kernel estimate and the discriminator starts with a strong capability to distinguish between patch distributions. Compared with state-of-the-art methods, our experiments show that MetaKernelGAN better estimates the magnitude and covariance of the kernel, leading to state-of-the-art blind SR results within a similar computational regime when combined with a non-blind SR model. Through supervised learning of an unsupervised learner, our method maintains the generalizability of the unsupervised learner, improves the optimization stability of kernel estimation, and hence image adaptation, and leads to a faster inference with a speedup between 14.24 to 102.1× over existing methods.0

Blind image super-resolution (SR) aims to recover high-resolution (HR) images from low-resolution (LR) inputs hindered by unknown degradation. Existing blind SR methods exploit computationally demanding explicit degradation estimators hinging on the availability of ground-truth information about the degradation process, thus introducing a severe limitation in real-world scenarios where this is inherently unattainable. Implicit degradation estimators avoid the need for ground truth but perform poorly. Our model reduces this performance gap with (i) a novel loss component to implicitly learn the degradation kernel from the LR input only, and (ii) a novel learnable Wiener filter module that exploits the learned degradation kernel to efficiently solve the deconvolution task via a closed-form solution formulated in the Fourier domain. Systematic experiments show that our proposed approach outperforms existing implicit blind SR methods (3dB PSNR gain and 8.5% SSIM improvement on average) and achieves comparable performance to explicit blind SR methods (0.6dB and 0.5% difference in PSNR and SSIM, respectively). Remarkably, these results are obtained using 33% and 71% less parameters than implicit and explicit methods.

… Finally, we apply the proposed degradation model to other networks … degradation model proposed in this paper can enhance the performance of remote sensing image super-resolution …

Deep learning-based methods have achieved significant successes on solving the blind super-resolution (BSR) problem. However, most of them request supervised pretraining on labelled datasets. This paper proposes an unsupervised kernel estimation model, named dynamic kernel prior (DKP), to realize an unsupervised and pretraining-free learning-based algorithm for solving the BSR problem. DKP can adaptively learn dynamic kernel priors to realize real-time kernel estimation, and thereby enables superior HR image restoration performances. This is achieved by a Markov chain Monte Carlo sampling process on random kernel distributions. The learned kernel prior is then assigned to optimize a blur kernel estimation network, which entails a network-based Langevin dynamic optimization strategy. These two techniques ensure the accuracy of the kernel estimation. DKP can be easily used to replace the kernel estimation models in the existing methods, such as Double-DIP and FKP-DIP, or be added to the off-the-shelf image restoration model, such as diffusion model. In this paper, we incorporate our DKP model with DIP and diffusion model, referring to DIP-DKP and Diff-DKP, for validations. Extensive simulations on Gaussian and motion kernel scenarios demonstrate that the proposed DKP model can significantly improve the kernel estimation with comparable runtime and memory usage, leading to state-of-the-art BSR results. The code is available at https://github.com/XYLGroup/DKP.

ABSTRACT Image super-resolution (SR) methods based on convolutional neural networks have become mainstream. However, unknown degradation in test images significantly affects SR performance. Existing blind SR methods struggle with inaccurate kernel estimation in explicit modeling and over-reliance on implicit degradation information. To address these issues, we propose the Degradation-Aware Dynamic Kernel with Dual-path Attention for Super-Resolution (DADKSR) network. DADKSR employs degradation representation learning to enable adaptive kernel estimation and uses a dual-path attention module to enhance spatial and channel features. An enhanced detail feature fusion (EDFF) module further enriches texture details, while the cycle correction learning strategy utilizes the SR results with the real degradation kernel in the low-resolution (LR) image to further correct the degradation kernel. Experimental results demonstrate that DADKSR improves kernel estimation accuracy, achieving an average PSNR improvement of 0.76 dB on public datasets with anisotropic Gaussian kernels, producing clearer structures and edges.

A variety of factors cause a reduction in remote sensing image resolution. Unlike super-resolution (SR) reconstruction methods with single degradation assumption, multi-degradation SR methods aim to learn the degradation kernel from low-resolution (LR) images and reconstruct high-resolution (HR) images more suitable for restoring the resolution of remote sensing images. However, existing multi-degradation SR methods only utilize the given LR images to learn the representation of the degradation kernel. The mismatches between the estimated degradation kernel and the real-world degradation kernel lead to a significant deterioration in performance of these methods. To address this issue, we design a reconstruction features-guided kernel correction SR network (RFKCNext) for multi-degradation SR reconstruction of remote sensing images. Specifically, the proposed network not only utilizes LR images to extract degradation kernel information but also employs features from SR images to correct the estimated degradation kernel, thereby enhancing the accuracy. RFKCNext utilizes the ConvNext Block (CNB) for global feature modeling. It employs CNB as fundamental units to construct the SR reconstruction subnetwork module (SRConvNext) and the reconstruction features-guided kernel correction network (RFGKCorrector). The SRConvNext reconstructs SR images based on the estimated degradation kernel. The RFGKCorrector corrects the estimated degradation kernel by reconstruction features from the generated SR images. The two networks iterate alternately, forming an end-to-end trainable network. More importantly, the SRConvNext utilizes the degradation kernel estimated by the RFGKCorrection for reconstruction, allowing the SRConvNext to perform well even if the degradation kernel deviates from the real-world scenario. In experimental terms, three levels of noise and five Gaussian blur kernels are considered on the NWPU-RESISC45 remote sensing image dataset for synthesizing degraded remote sensing images to train and test. Compared to existing super-resolution methods, the experimental results demonstrate that our proposed approach achieves significant reconstruction advantages in both quantitative and qualitative evaluations. Additionally, the UCMERCED remote sensing dataset and the real-world remote sensing image dataset provided by the “Tianzhi Cup” Artificial Intelligence Challenge are utilized for further testing. Extensive experiments show that our method delivers more visually plausible results, demonstrating the potential of real-world application.

… imaging physical kernel is proposed. The imaging degradation factors of the infrared images are analyzed, and the modulation transfer function of the infrared thermal imaging system is …

… of producing a high-quality HR image from the LR input. … for harnessing the acquired image-specific SR kernel. KGSR … both the image-specific SR kernel and high-quality HR images. …

Various super-resolution (SR) kernels in the degradation model deteriorate the performance of the SR algorithms, showing unpleasant artifacts in the output images. Hence, SR kernel estimation has been studied to improve the SR performance in several ways for more than a decade. In particular, a conventional research named KernelGAN has recently been proposed. To estimate the SR kernel from a single image, KernelGAN introduces generative adversarial networks(GANs) that utilize the recurrence of similar structures across scales. Subsequently, an enhanced version of KernelGAN, named E-KernelGAN, was proposed to consider image sharpness and edge thickness. Although it is stable compared to the earlier method, it still encounters challenges in estimating sizable and anisotropic kernels because the structural information of an input image is not sufficiently considered. In this paper, we propose a kernel estimation algorithm called Total Variation Guided KernelGAN (TVG-KernelGAN), which efficiently enables networks to focus on the structural information of an input image. The experimental results show that the proposed algorithm accurately and stably estimates kernels, particularly sizable and anisotropic kernels, both qualitatively and quantitatively. In addition, we compared the results of the non-blind SR methods, using SR kernel estimation techniques. The results indicate that the performance of the SR algorithms was improved using our proposed method.

The goal of blind image super-resolution (BISR) is to recover the corresponding high-resolution image from a given low-resolution image with unknown degradation. Prior related research has primarily focused effectively on utilizing the kernel as prior knowledge to recover the high-frequency components of image. However, they overlooked the function of structural prior information within the same image, which resulted in unsatisfactory recovery performance for textures with strong self-similarity. To address this issue, we propose a two stage blind super-resolution network that is based on kernel estimation strategy and is capable of integrating structural texture as prior knowledge. In the first stage, we utilize a dynamic kernel estimator to achieve degradation presentation embedding. Then, we propose a triple path attention groups consists of triple path attention blocks and a global feature fusion block to extract structural prior information to assist the recovery of details within images. The quantitative and qualitative results on standard benchmarks with various degradation settings, including Gaussian8 and DIV2KRK, validate that our proposed method outperforms the state-of-the-art methods in terms of fidelity and recovery of clear details. The relevant code is made available on this link as open source.

Existing Blind image Super-Resolution (BSR) methods focus on estimating either kernel or degradation infor-mation, but have long overlooked the essential content details. In this paper, we propose a novel BSR approach, Content-aware Degradation-driven Transformer (CDFormer), to capture both degradation and content rep-resentations. However, low-resolution images cannot pro-vide enough content details, and thus we introduce a diffusion-based module CD Former dif f to first learn Con-tent Degradation Prior (CDP) in both low- and high-resolution images, and then approximate the real distribution given only low-resolution information. Moreover, we apply an adaptive SR network CDFormersR that effectively utilizes CDP to refine features. Compared to previous diffusion-based SR methods, we treat the diffusion model as an estimator that can overcome the limitations of expensive sampling time and excessive diversity. Experiments show that CDFormer can outperform existing methods, establishing a new state-of-the-art performance on various bench-marks under blind settings. Codes and models will be avail-able at https://github.com/I2-Multimedia-Lab/CDFormer.

Recent blind super-resolution (BSR) methods are explored to handle unknown degradations and achieve impressive performance. However, the prevailing assumption in most BSR methods is the spatial invariance of degradation kernels across the entire image, which leads to significant performance declines when faced with spatially variant degradations caused by object motion or defocusing. Additionally, these methods do not account for the human visual system's tendency to focus differently on areas of varying perceptual difficulty, as they uniformly process each pixel during reconstruction. To cope with these issues, we propose a difficulty-guided variant degradation learning network for BSR, named difficulty-guided degradation learning (DDL)-BSR, which explores the relationship between reconstruction difficulty and degradation estimation. Accordingly, the proposed DDL-BSR consists of three customized networks: reconstruction difficulty prediction (RDP), space-variant degradation estimation (SDE), and degradation and difficulty-informed reconstruction (DDR). Specifically, RDP learns the reconstruction difficulty with the proposed reconstruction-distance supervision. Then, SDE is designed to estimate space-variant degradation kernels according to the difficulty map. Finally, both degradation kernels and reconstruction difficulty are fed into DDR, which takes into account such two prior knowledge information to guide super-resolution (SR). Experimental analysis on various synthetic datasets demonstrates that DDL-BSR invariably surpasses state-of-the-art (SOTA) methods, producing SR images with enhanced realism and texture quality. Code is available at https://github.com/JiaWang0704/DDL-BSR.

… Real-world super-resolution via kernel estimation and noise injection. In Proceedings of the IEEE/CVF … Given a degradation, we use the classical image degradation model [17…

Existing methods for single image super-resolution (SISR) model the blur kernel as spatially invariant across the entire image, and are susceptible to the adverse effects of textureless patches. To achieve improved results, adaptive estimation of the degradation kernel is necessary. We explore the synergy of joint global and local degradation modeling for spatially adaptive blind SISR. Our model, named spatially adaptive network for blind super-resolution (SASR), employs a simple encoder to estimate global degradation representations and a decoder to extract local degradation. These two representations are fused with a cross-attention mechanism and applied using spatially adaptive filtering to enhance the local image detail. Specifically, SASR contains two novel features: (1) a non-local degradation modeling with contrastive learning to learn global and local degradation representations, and (2) a non-local spatially adaptive filtering module (SAFM) that incorporates the global degradation and spatial-detail factors to preserve and enhance local details. We demonstrate that SASR can efficiently estimate degradation representations and handle multiple types of degradation. The local representations avoid the detrimental effect of estimating the entire super-resolved image with only one kernel through locally adaptive adjustments. Extensive experiments are performed to quantitatively and qualitatively demonstrate that SASR not only performs favorably for degradation estimation but also leads to state-of-the-art blind SISR performance when compared to alternative approaches.

Blind super-resolution (BlindSR) has recently attracted attention in the field of remote sensing. Due to the lack of paired data, most works assume that the acquired remote sensing images are high-resolution (HR) and use predefined degradation models to synthesize low-resolution (LR) images for training and evaluation. However, these acquired remote sensing images are often degraded by various factors, which still require super-resolution (SR) reconstruction to meet practical needs. Using them as ground-truth (GT) images will limit the model’s ability to restore fine details, resulting in blurry and noisy reconstructions. To overcome these limitations, we propose an unsupervised degradation-aware network which transforms natural images into the degraded domain as real-world remote sensing images. It uses natural images containing rich texture information as a reference for fine-grained restoration of the network, enabling the network to produce clearer reconstructions. Furthermore, we discovered the remarkable capability of the patchwise discriminator to perceive the degradation type of different regions within the acquired remote sensing image. Inspired by this finding, we design a novel degradation representation module (DRM) that can estimate the degradation information from LR images and guide the network to perform adaptive restoration. Comprehensive experimental results demonstrate that our proposed unsupervised blind super-resolution framework achieves state-of-the-art (SOTA) restoration performance. Our code and pretrained models have been uploaded to GitHub (https://github.com/55Dupup/UDASR) for validation.

Blind Super-Resolution (BlindSR) aims to reconstruct high-resolution (HR) images from low-resolution (LR) images without prior knowledge of the image degradation process. This is a challenging problem in real-world applications, where the degradation can be complex and unknown. Recent unsupervised learning-based BlindSR methods can estimate the image degradation in an unsupervised manner, but they suffer from limited adaptability to different types and intensities of degradation. They tend to capture the average level of degradation across all training samples, resulting in over-smoothing or over-sharpening effects for some images. As a result, the final reconstruction may exhibit the mean effect. Moreover, existing synthetic datasets do not reflect the real-world degradation scenarios, making it difficult to evaluate the performance of BlindSR methods. To address these issues, we propose a novel Degradation Intensity Estimation Module (DIEM) method, which can estimate the pixel-level degradation information of the input image more specifically and use it to guide image reconstruction. Furthermore, we construct a benchmark dataset under real scenarios, which is closer to the real-world BlindSR problem than existing synthetic datasets, and can provide a more reasonable evaluation of BlindSR methods. Extensive experimental results demonstrate that our DIEM-guided BlindSR method can achieve state-of-the-art image reconstruction results. Our code and pre-trained models have been uploaded to GitHub for validation.

… based self-supervised blind image super-resolution method (Self-SR), which models a joint optimization problem about the blur … down-sampling or assume the blur kernel is known. The …

… We propose an effective self-supervised learning module to estimate blur kernels and … apply the estimated blur kernel \(N_k(L)\) to L and use the downsampling operation to generate a …

… Existing image super-resolution (SR) methods often rely on implicit or explicit estimation of … components as supervised signals for self-supervised training. This design enables our …

… poor generalization performance and cannot handle unknown degradation. Additionally, … image blind super-resolution algorithm. Specifically, we first use a blur kernel estimation …

Previous studies in blind super-resolution (BSR) have primarily concentrated on estimating degradation kernels directly from low-resolution (LR) inputs to enhance super-resolution. However, these degradation kernels, which model the transition from a high-resolution (HR) image to its LR version, should account for not only the degradation process but also the downscaling factor. Applying the same degradation kernel across varying super-resolution scales may be impractical. Our research acknowledges degradation kernels and scaling factors as pivotal elements for the BSR task and introduces a novel strategy that utilizes HR images as references to establish scale-aware degradation kernels. By employing content-irrelevant HR reference images alongside the target LR image, our model adaptively discerns the degradation process. It is then applied to generate additional LR-HR pairs through down-sampling the HR reference images, which are keys to improving the SR performance. Our reference-based training procedure is applicable to proficiently trained blind SR models and zero-shot blind SR methods, consistently outperforming previous methods in both scenarios. This dual consideration of blur kernels and scaling factors, coupled with the use of a reference image, contributes to the effectiveness of our approach in blind super-resolution tasks.

Image super-resolution (SR) usually synthesizes degraded low-resolution images with a predefined degradation model for training. Existing SR methods inevitably perform poorly when the true degradation does not follow the predefined degradation, especially in the case of the real world. To tackle this robustness issue, we propose a cascaded degradation-aware blind super-resolution network (CDASRN), which not only eliminates the influence of noise on blur kernel estimation but also can estimate the spatially varying blur kernel. With the addition of contrastive learning, our CDASRN can further distinguish the differences between local blur kernels, greatly improving its practicality. Experiments in various settings show that CDASRN outperforms state-of-the-art methods on both heavily degraded synthetic datasets and real-world datasets.

Convolutional neural network-based super-resolution (SR) methods have achieved significant success on ideal, predefined downsampling (bicubic) kernels. However, these algorithms struggle with unknown degradations in real-world data, which often follow a spatially variant and unknown distribution. Recently proposed blind SR studies address this issue by estimating degradation kernels, but their results often exhibit artifacts and detail deformation due to redundant information being considered in kernel estimation. Additionally, effective merging of the estimated kernels into the feature space of the SR network is challenging. To overcome these issues, we introduce a novel network, KernFusNet which simultaneously learns the kernel degradation and the relevant content information to adapt to the blur characteristics in real-world images. Specifically, KernFusNet comprises two components: an Implicit Kernel Estimation (IKE) module and a Kernel-Prior Oriented Detail Fusion (KPDF) module. The IKE module estimates the degradation kernel from low-resolution contexts, while the KPDF module effectively merge the relevant information on the basis of the learned degradations in both high-resolution and low-resolution spaces. Comprehensive experiments on the real-world and synthetic datasets show that our network achieves state-of-the-art performance for blind SR.

… , blind super-resolution is a key area focused on generating high-resolution images with enhanced visual quality from low-resolution counterparts affected by indeterminate degradation …

Blind image super-resolution (SR) aims to recover a high-resolution (HR) image from its low-resolution (LR) counterpart under the assumption of unknown degradations. Many existing blind SR methods rely on supervising ground-truth kernels referred to as explicit degradation estimators. However, it is very challenging to obtain the ground-truths for different degradations kernels. Moreover, most of these methods rely on heavy backbone networks, which demand extensive computational resources. Implicit degradation estimators do not require the availability of ground truth kernels, but they see a significant performance gap with the explicit degradation estimators due to such missing information. We present a novel approach that significantly narrows such a gap by means of a lightweight architecture that implicitly learns the degradation kernel with the help of a novel loss component. The kernel is exploited by a learnable Wiener filter that performs efficient deconvolution in the Fourier domain by deriving a closed-form solution. Inspired by prompt-based learning, we also propose a novel degradation-conditioned prompt layer that exploits the estimated kernel to drive the focus on the discriminative contextual information that guides the reconstruction process in recovering the latent HR image. Extensive experiments under different degradation settings demonstrate that our model, named PL-IDENet, yields PSNR and SSIM improvements of more than $0.4dB$ and 1.3%, and $1.4dB$ and 4.8% to the best implicit and explicit blind-SR method, respectively. These results are achieved while maintaining a substantially lower number of parameters/FLOPs (i.e., 25% and 68% fewer parameters than best implicit and explicit methods, respectively).

… (SR) methods are performed under a known or specific degradation kernel. However, the … image blind super-resolution model (Med-BSR) based on an improved degradation process to …

Magnetic Resonance Spectroscopic Imaging (MRSI) is a non-invasive imaging technique for studying metabolism and has become a crucial tool for understanding neurological diseases, cancers and diabetes. High spatial resolution MRSI is needed to characterize lesions, but in practice MRSI is acquired at low resolution due to time and sensitivity restrictions caused by the low metabolite concentrations. Therefore, there is an imperative need for a post-processing approach to generate high-resolution MRSI from low-resolution data that can be acquired fast and with high sensitivity. Deep learning-based super-resolution methods provided promising results for improving the spatial resolution of MRSI, but they still have limited capability to generate accurate and high-quality images. Recently, diffusion models have demonstrated superior learning capability than other generative models in various tasks, but sampling from diffusion models requires iterating through a large number of diffusion steps, which is time-consuming. This work introduces a Flow-based Truncated Denoising Diffusion Model (FTDDM) for super-resolution MRSI, which shortens the diffusion process by truncating the diffusion chain, and the truncated steps are estimated using a normalizing flow-based network. The network is conditioned on upscaling factors to enable multi-scale super-resolution. To train and evaluate the deep learning models, we developed a 1H-MRSI dataset acquired from 25 high-grade glioma patients. We demonstrate that FTDDM outperforms existing generative models while speeding up the sampling process by over 9-fold compared to the baseline diffusion model. Neuroradiologists' evaluations confirmed the clinical advantages of our method, which also supports uncertainty estimation and sharpness adjustment, extending its potential clinical applications.

In recent years, Vision Transformer-based approaches for low-level vision tasks have achieved widespread success. Unlike CNN-based models, Transformers are more adept at capturing long-range dependencies, enabling the reconstruction of images utilizing non-local information. In the domain of super-resolution, Swin-transformer-based models have become mainstream due to their capability of global spatial information modeling and their shifting-window attention mechanism that facilitates the interchange of information between different windows. Many researchers have enhanced model performance by expanding the receptive fields or designing meticulous networks, yielding commendable results. However, we observed that it is a general phenomenon for the feature map intensity to be abruptly suppressed to small values towards the network's end. This implies an information bottleneck and a diminishment of spatial information, implicitly limiting the model's potential. To address this, we propose the Dense-residual-connected Transformer (DRCT), aimed at mitigating the loss of spatial information and stabilizing the information flow through dense-residual connections between layers, thereby unleashing the model's potential and saving the model away from information bottleneck. Experiment results indicate that our approach surpasses state-of-the-art methods on benchmark datasets and performs commendably at the NTIRE-2024 Image Super-Resolution (x4) Challenge. Our source code is available at https://github.com/ming053l/DRCT

Recently, Transformer-based methods have achieved impressive results in single image super-resolution (SISR). However, the lack of locality mechanism and high complexity limit their application in the field of super-resolution (SR). To solve these problems, we propose a new method, Efficient Mixed Transformer (EMT) in this study. Specifically, we propose the Mixed Transformer Block (MTB), consisting of multiple consecutive transformer layers, in some of which the Pixel Mixer (PM) is used to replace the Self-Attention (SA). PM can enhance the local knowledge aggregation with pixel shifting operations. At the same time, no additional complexity is introduced as PM has no parameters and floating-point operations. Moreover, we employ striped window for SA (SWSA) to gain an efficient global dependency modelling by utilizing image anisotropy. Experimental results show that EMT outperforms the existing methods on benchmark dataset and achieved state-of-the-art performance. The Code is available at https://github.com/Fried-Rice-Lab/FriedRiceLab.

Transformer-based models have achieved remarkable results in low-level vision tasks including image super-resolution (SR). However, early Transformer-based approaches that rely on self-attention within non-overlapping windows encounter challenges in acquiring global information. To activate more input pixels globally, hybrid attention models have been proposed. Moreover, training by solely minimizing pixel-wise RGB losses, such as L1, have been found inadequate for capturing essential high-frequency details. This paper presents two contributions: i) We introduce convolutional non-local sparse attention (NLSA) blocks to extend the hybrid transformer architecture in order to further enhance its receptive field. ii) We employ wavelet losses to train Transformer models to improve quantitative and subjective performance. While wavelet losses have been explored previously, showing their power in training Transformer-based SR models is novel. Our experimental results demonstrate that the proposed model provides state-of-the-art PSNR results as well as superior visual performance across various benchmark datasets.

Blind single image super-resolution (SISR) is a challenging task in image processing due to the ill-posed nature of the inverse problem. Complex degradations present in real life images make it difficult to solve this problem using naïve deep learning approaches, where models are often trained on synthetically generated image pairs. Most of the effort so far has been focused on solving the inverse problem under some constraints, such as for a limited space of blur kernels and/or assuming noise-free input images. Yet, there is a gap in the literature to provide a well-generalized deep learning-based solution that performs well on images with unknown and highly complex degradations. In this paper, we propose IKR-Net (Iterative Kernel Reconstruction Network) for blind SISR. In the proposed approach, kernel and noise estimation and high-resolution image reconstruction are carried out iteratively using dedicated deep models. The iterative refinement provides significant improvement in both the reconstructed image and the estimated blur kernel even for noisy inputs. IKR-Net provides a generalized solution that can handle any type of blur and level of noise in the input low-resolution image. IKR-Net achieves state-of-the-art results in blind SISR, especially for noisy images with motion blur.

In recent years, deep neural networks, including Convolutional Neural Networks, Transformers, and State Space Models, have achieved significant progress in Remote Sensing Image (RSI) Super-Resolution (SR). However, existing SR methods typically overlook the complementary relationship between global and local dependencies. These methods either focus on capturing local information or prioritize global information, which results in models that are unable to effectively capture both global and local features simultaneously. Moreover, their computational cost becomes prohibitive when applied to large-scale RSIs. To address these challenges, we introduce the novel application of Receptance Weighted Key Value (RWKV) to RSI-SR, which captures long-range dependencies with linear complexity. To simultaneously model global and local features, we propose the Global-Detail dual-branch structure, GDSR, which performs SR by paralleling RWKV and convolutional operations to handle large-scale RSIs. Furthermore, we introduce the Global-Detail Reconstruction Module (GDRM) as an intermediary between the two branches to bridge their complementary roles. In addition, we propose the Dual-Group Multi-Scale Wavelet Loss, a wavelet-domain constraint mechanism via dual-group subband strategy and cross-resolution frequency alignment for enhanced reconstruction fidelity in RSI-SR. Extensive experiments under two degradation methods on several benchmarks, including AID, UCMerced, and RSSRD-QH, demonstrate that GSDR outperforms the state-of-the-art Transformer-based method HAT by an average of 0.09 dB in PSNR, while using only 63% of its parameters and 51% of its FLOPs, achieving an inference speed 3.2 times faster.

Recently, the methods based on implicit neural representations have shown excellent capabilities for arbitrary-scale super-resolution (ASSR). Although these methods represent the features of an image by generating latent codes, these latent codes are difficult to adapt for different magnification factors of super-resolution, which seriously affects their performance. Addressing this, we design Multi-Scale Implicit Transformer (MSIT), consisting of an Multi-scale Neural Operator (MSNO) and Multi-Scale Self-Attention (MSSA). Among them, MSNO obtains multi-scale latent codes through feature enhancement, multi-scale characteristics extraction, and multi-scale characteristics merging. MSSA further enhances the multi-scale characteristics of latent codes, resulting in better performance. Furthermore, to improve the performance of network, we propose the Re-Interaction Module (RIM) combined with the cumulative training strategy to improve the diversity of learned information for the network. We have systematically introduced multi-scale characteristics for the first time in ASSR, extensive experiments are performed to validate the effectiveness of MSIT, and our method achieves state-of-the-art performance in arbitrary super-resolution tasks.

Deep convolutional neural networks can use hierarchical information to progressively extract structural information to recover high-quality images. However, preserving the effectiveness of the obtained structural information is important in image super-resolution. In this paper, we propose a cosine network for image super-resolution (CSRNet) by improving a network architecture and optimizing the training strategy. To extract complementary homologous structural information, odd and even heterogeneous blocks are designed to enlarge the architectural differences and improve the performance of image super-resolution. Combining linear and non-linear structural information can overcome the drawback of homologous information and enhance the robustness of the obtained structural information in image super-resolution. Taking into account the local minimum of gradient descent, a cosine annealing mechanism is used to optimize the training procedure by performing warm restarts and adjusting the learning rate. Experimental results illustrate that the proposed CSRNet is competitive with state-of-the-art methods in image super-resolution.

Physically Based Rendering (PBR) materials are typically characterized by multiple 2D texture maps such as basecolor, normal, metallic, and roughness which encode spatially-varying bi-directional reflectance distribution function (SVBRDF) parameters to model surface reflectance properties and microfacet interactions. Upscaling SVBRDF material is valuable for modern 3D graphics applications. However, existing Single Image Super-Resolution (SISR) methods struggle with cross-map inconsistency, inadequate modeling of modality-specific features, and limited generalization due to data distribution shifts. In this work, we propose Multi-modal Upscaling Joint Inference via Cross-map Attention (MUJICA), a flexible adapter that reforms pre-trained Swin-transformer-based SISR models for PBR material super-resolution. MUJICA is seamlessly attached after the pre-trained and frozen SISR backbone. It leverages cross-map attention to fuse features while preserving remarkable reconstruction ability of the pre-trained SISR model. Applied to SISR models such as SwinIR, DRCT, and HMANet, MUJICA improves PSNR, SSIM, and LPIPS scores while preserving cross-map consistency. Experiments demonstrate that MUJICA enables efficient training even with limited resources and delivers state-of-the-art performance on PBR material datasets.

This paper reviews the NTIRE 2026 challenge on efficient single-image super-resolution with a focus on the proposed solutions and results. The aim of this challenge is to devise a network that reduces one or several aspects, such as runtime, parameters, and FLOPs, while maintaining PSNR of around 26.90 dB on the DIV2K_LSDIR_valid dataset, and 26.99 dB on the DIV2K_LSDIR_test dataset. The challenge had 95 registered participants, and 15 teams made valid submissions. They gauge the state-of-the-art results for efficient single-image super-resolution.

This paper provides a comprehensive review of the NTIRE 2024 challenge, focusing on efficient single-image super-resolution (ESR) solutions and their outcomes. The task of this challenge is to super-resolve an input image with a magnification factor of x4 based on pairs of low and corresponding high-resolution images. The primary objective is to develop networks that optimize various aspects such as runtime, parameters, and FLOPs, while still maintaining a peak signal-to-noise ratio (PSNR) of approximately 26.90 dB on the DIV2K_LSDIR_valid dataset and 26.99 dB on the DIV2K_LSDIR_test dataset. In addition, this challenge has 4 tracks including the main track (overall performance), sub-track 1 (runtime), sub-track 2 (FLOPs), and sub-track 3 (parameters). In the main track, all three metrics (ie runtime, FLOPs, and parameter count) were considered. The ranking of the main track is calculated based on a weighted sum-up of the scores of all other sub-tracks. In sub-track 1, the practical runtime performance of the submissions was evaluated, and the corresponding score was used to determine the ranking. In sub-track 2, the number of FLOPs was considered. The score calculated based on the corresponding FLOPs was used to determine the ranking. In sub-track 3, the number of parameters was considered. The score calculated based on the corresponding parameters was used to determine the ranking. RLFN is set as the baseline for efficiency measurement. The challenge had 262 registered participants, and 34 teams made valid submissions. They gauge the state-of-the-art in efficient single-image super-resolution. To facilitate the reproducibility of the challenge and enable other researchers to build upon these findings, the code and the pre-trained model of validated solutions are made publicly available at https://github.com/Amazingren/NTIRE2024_ESR/.

This paper presents a comprehensive review of the NTIRE 2025 Challenge on Single-Image Efficient Super-Resolution (ESR). The challenge aimed to advance the development of deep models that optimize key computational metrics, i.e., runtime, parameters, and FLOPs, while achieving a PSNR of at least 26.90 dB on the $\operatorname{DIV2K\_LSDIR\_valid}$ dataset and 26.99 dB on the $\operatorname{DIV2K\_LSDIR\_test}$ dataset. A robust participation saw \textbf{244} registered entrants, with \textbf{43} teams submitting valid entries. This report meticulously analyzes these methods and results, emphasizing groundbreaking advancements in state-of-the-art single-image ESR techniques. The analysis highlights innovative approaches and establishes benchmarks for future research in the field.

Deep convolutional neural networks can extract more accurate structural information via deep architectures to obtain good performance in image super-resolution. However, it is not easy to find effect of important layers in a single network architecture to decrease performance of super-resolution. In this paper, we design a tree-guided CNN for image super-resolution (TSRNet). It uses a tree architecture to guide a deep network to enhance effect of key nodes to amplify the relation of hierarchical information for improving the ability of recovering images. To prevent insufficiency of the obtained structural information, cosine transform techniques in the TSRNet are used to extract cross-domain information to improve the performance of image super-resolution. Adaptive Nesterov momentum optimizer (Adan) is applied to optimize parameters to boost effectiveness of training a super-resolution model. Extended experiments can verify superiority of the proposed TSRNet for restoring high-quality images. Its code can be obtained at https://github.com/hellloxiaotian/TSRNet.

Transformers have achieved remarkable results in single-image super-resolution (SR). However, the challenge of balancing model performance and complexity has hindered their application in lightweight SR (LSR). To tackle this challenge, we propose an efficient striped window transformer (ESWT). We revisit the normalization layer in the transformer and design a concise and efficient transformer structure to build the ESWT. Furthermore, we introduce a striped window mechanism to model long-term dependencies more efficiently. To fully exploit the potential of the ESWT, we propose a novel flexible window training strategy that can improve the performance of the ESWT without additional cost. Extensive experiments show that ESWT outperforms state-of-the-art LSR transformers, and achieves a better trade-off between model performance and complexity. The ESWT requires fewer parameters, incurs faster inference, smaller FLOPs, and less memory consumption, making it a promising solution for LSR.

This work tackles the fidelity objective in the perceptual super-resolution~(SR). Specifically, we address the shortcomings of pixel-level $L_\text{p}$ loss ($\mathcal{L}_\text{pix}$) in the GAN-based SR framework. Since $L_\text{pix}$ is known to have a trade-off relationship against perceptual quality, prior methods often multiply a small scale factor or utilize low-pass filters. However, this work shows that these circumventions fail to address the fundamental factor that induces blurring. Accordingly, we focus on two points: 1) precisely discriminating the subcomponent of $L_\text{pix}$ that contributes to blurring, and 2) only guiding based on the factor that is free from this trade-off relationship. We show that they can be achieved in a surprisingly simple manner, with an Auto-Encoder (AE) pretrained with $L_\text{pix}$. Accordingly, we propose the Auto-Encoded Supervision for Optimal Penalization loss ($L_\text{AESOP}$), a novel loss function that measures distance in the AE space, instead of the raw pixel space. Note that the AE space indicates the space after the decoder, not the bottleneck. By simply substituting $L_\text{pix}$ with $L_\text{AESOP}$, we can provide effective reconstruction guidance without compromising perceptual quality. Designed for simplicity, our method enables easy integration into existing SR frameworks. Experimental results verify that AESOP can lead to favorable results in the perceptual SR task.

We present a novel method for diffusion-guided frameworks for view-consistent super-resolution (SR) in neural rendering. Our approach leverages existing 2D SR models in conjunction with advanced techniques such as Variational Score Distilling (VSD) and a LoRA fine-tuning helper, with spatial training to significantly boost the quality and consistency of upscaled 2D images compared to the previous methods in the literature, such as Renoised Score Distillation (RSD) proposed in DiSR-NeRF (1), or SDS proposed in DreamFusion. The VSD score facilitates precise fine-tuning of SR models, resulting in high-quality, view-consistent images. To address the common challenge of inconsistencies among independent SR 2D images, we integrate Iterative 3D Synchronization (I3DS) from the DiSR-NeRF framework. Our quantitative benchmarks and qualitative results on the LLFF dataset demonstrate the superior performance of our system compared to existing methods such as DiSR-NeRF.

This study demonstrates that a transformer-based neural operator (TNO) can perform zero-shot super-resolution of two-dimensional temperature fields near the ground in urban areas. During training, super-resolution is performed from a horizontal resolution of 100 m to 20 m, while during testing, it is performed from 100 m to a finer resolution of 5 m. This setting is referred to as zero-shot, since no data with the target 5 m resolution are included in the training dataset. The 20 m and 5 m resolution data were independently obtained by dynamically downscaling the 100 m data using a physics-based micrometeorology model that resolves buildings. Compared to a convolutional neural network, the TNO more accurately reproduces temperature distributions at 5 m resolution and reduces test errors by approximately 33%. Furthermore, the TNO successfully performs zero-shot super-resolution even when trained with unstructured data, in which grid points are randomly arranged. These results suggest that the TNO recognizes building shapes independently of grid point locations and adaptively infers the temperature fields induced by buildings.

Diffusion generative models have achieved remarkable success in generating images with a fixed resolution. However, existing models have limited ability to generalize to different resolutions when training data at those resolutions are not available. Leveraging techniques from operator learning, we present a novel deep-learning architecture, Dual-FNO UNet (DFU), which approximates the score operator by combining both spatial and spectral information at multiple resolutions. Comparisons of DFU to baselines demonstrate its scalability: 1) simultaneously training on multiple resolutions improves FID over training at any single fixed resolution; 2) DFU generalizes beyond its training resolutions, allowing for coherent, high-fidelity generation at higher-resolutions with the same model, i.e. zero-shot super-resolution image-generation; 3) we propose a fine-tuning strategy to further enhance the zero-shot super-resolution image-generation capability of our model, leading to a FID of 11.3 at 1.66 times the maximum training resolution on FFHQ, which no other method can come close to achieving.

Analysis of compressible turbulent flows is essential for applications related to propulsion, energy generation, and the environment. Here, we present BLASTNet 2.0, a 2.2 TB network-of-datasets containing 744 full-domain samples from 34 high-fidelity direct numerical simulations, which addresses the current limited availability of 3D high-fidelity reacting and non-reacting compressible turbulent flow simulation data. With this data, we benchmark a total of 49 variations of five deep learning approaches for 3D super-resolution - which can be applied for improving scientific imaging, simulations, turbulence models, as well as in computer vision applications. We perform neural scaling analysis on these models to examine the performance of different machine learning (ML) approaches, including two scientific ML techniques. We demonstrate that (i) predictive performance can scale with model size and cost, (ii) architecture matters significantly, especially for smaller models, and (iii) the benefits of physics-based losses can persist with increasing model size. The outcomes of this benchmark study are anticipated to offer insights that can aid the design of 3D super-resolution models, especially for turbulence models, while this data is expected to foster ML methods for a broad range of flow physics applications. This data is publicly available with download links and browsing tools consolidated at https://blastnet.github.io.

Digital zoom on smartphones relies on learning-based super-resolution (SR) models that operate on RAW sensor images, but obtaining sensor-specific training data is challenging due to the lack of ground-truth images. Synthetic data generation via ``unprocessing'' pipelines offers a potential solution by simulating the degradations that transform high-resolution (HR) images into their low-resolution (LR) counterparts. However, these pipelines can introduce domain gaps due to incomplete or unrealistic degradation modeling. In this paper, we demonstrate that principled and carefully designed degradation modeling can enhance SR performance in real-world conditions. Instead of relying on generic priors for camera blur and noise, we model device-specific degradations through calibration and unprocess publicly available rendered images into the RAW domain of different smartphones. Using these image pairs, we train a single-image RAW-to-RGB SR model and evaluate it on real data from a held-out device. Our experiments show that accurate degradation modeling leads to noticeable improvements, with our SR model outperforming baselines trained on large pools of arbitrarily chosen degradations.

This paper focuses on the dataset-free Blind Image Super-Resolution (BISR). Unlike existing dataset-free BISR methods that focus on obtaining a degradation kernel for the entire image, we are the first to explicitly design a spatially-variant degradation model for each pixel. Our method also benefits from having a significantly smaller number of learnable parameters compared to data-driven spatially-variant BISR methods. Concretely, each pixel's degradation kernel is expressed as a linear combination of a learnable dictionary composed of a small number of spatially-variant atom kernels. The coefficient matrices of the atom degradation kernels are derived using membership functions of fuzzy set theory. We construct a novel Probabilistic BISR model with tailored likelihood function and prior terms. Subsequently, we employ the Monte Carlo EM algorithm to infer the degradation kernels for each pixel. Our method achieves a significant improvement over other state-of-the-art BISR methods, with an average improvement of 1 dB (2x).Code will be released at https://github.com/shaojieguoECNU/SVDSR.

Recent advancements in light field super-resolution (SR) have yielded impressive results. In practice, however, many existing methods are limited by assuming fixed degradation models, such as bicubic downsampling, which hinders their robustness in real-world scenarios with complex degradations. To address this limitation, we present LF-DEST, an effective blind Light Field SR method that incorporates explicit Degradation Estimation to handle various degradation types. LF-DEST consists of two primary components: degradation estimation and light field restoration. The former concurrently estimates blur kernels and noise maps from low-resolution degraded light fields, while the latter generates super-resolved light fields based on the estimated degradations. Notably, we introduce a modulated and selective fusion module that intelligently combines degradation representations with image information, allowing for effective handling of diverse degradation types. We conduct extensive experiments on benchmark datasets, demonstrating that LF-DEST achieves superior performance across a variety of degradation scenarios in light field SR.

Unsupervised real-world super-resolution (SR) faces critical challenges due to the complex, unknown degradation distributions in practical scenarios. Existing methods struggle to generalize from synthetic low-resolution (LR) and high-resolution (HR) image pairs to real-world data due to a significant domain gap. In this paper, we propose an unsupervised real-world SR method based on rectified flow to effectively capture and model real-world degradation, synthesizing LR-HR training pairs with realistic degradation. Specifically, given unpaired LR and HR images, we propose a novel Rectified Flow Degradation Module (RFDM) that introduces degradation-transformed LR (DT-LR) images as intermediaries. By modeling the degradation trajectory in a continuous and invertible manner, RFDM better captures real-world degradation and enhances the realism of generated LR images. Additionally, we propose a Fourier Prior Guided Degradation Module (FGDM) that leverages structural information embedded in Fourier phase components to ensure more precise modeling of real-world degradation. Finally, the LR images are processed by both FGDM and RFDM, producing final synthetic LR images with real-world degradation. The synthetic LR images are paired with the given HR images to train the off-the-shelf SR networks. Extensive experiments on real-world datasets demonstrate that our method significantly enhances the performance of existing SR approaches in real-world scenarios.

Implicit degradation modeling-based blind super-resolution (SR) has attracted more increasing attention in the community due to its excellent generalization to complex degradation scenarios and wide application range. How to extract more discriminative degradation representations and fully adapt them to specific image features is the key to this task. In this paper, we propose a new Content-decoupled Contrastive Learning-based blind image super-resolution (CdCL) framework following the typical blind SR pipeline. This framework introduces negative-free contrastive learning technique for the first time to model the implicit degradation representation, in which a new cyclic shift sampling strategy is designed to ensure decoupling between content features and degradation features from the data perspective, thereby improving the purity and discriminability of the learned implicit degradation space. In addition, we propose a detail-aware implicit degradation adapting module that can better adapt degradation representations to specific LR features by enhancing the basic adaptation unit's perception of image details, significantly reducing the overall SR model complexity. Extensive experiments on synthetic and real data show that our method achieves highly competitive quantitative and qualitative results in various degradation settings while obviously reducing parameters and computational costs, validating the feasibility of designing practical and lightweight blind SR tools.

This paper reviews the NTIRE 2025 RAW Image Restoration and Super-Resolution Challenge, highlighting the proposed solutions and results. New methods for RAW Restoration and Super-Resolution could be essential in modern Image Signal Processing (ISP) pipelines, however, this problem is not as explored as in the RGB domain. The goal of this challenge is two fold, (i) restore RAW images with blur and noise degradations, (ii) upscale RAW Bayer images by 2x, considering unknown noise and blur. In the challenge, a total of 230 participants registered, and 45 submitted results during thee challenge period. This report presents the current state-of-the-art in RAW Restoration.

Prior methodologies have disregarded the diversities among distinct degradation types during image reconstruction, employing a uniform network model to handle multiple deteriorations. Nevertheless, we discover that prevalent degradation modalities, including sampling, blurring, and noise, can be roughly categorized into two classes. We classify the first class as spatial-agnostic dominant degradations, less affected by regional changes in image space, such as downsampling and noise degradation. The second class degradation type is intimately associated with the spatial position of the image, such as blurring, and we identify them as spatial-specific dominant degradations. We introduce a dynamic filter network integrating global and local branches to address these two degradation types. This network can greatly alleviate the practical degradation problem. Specifically, the global dynamic filtering layer can perceive the spatial-agnostic dominant degradation in different images by applying weights generated by the attention mechanism to multiple parallel standard convolution kernels, enhancing the network's representation ability. Meanwhile, the local dynamic filtering layer converts feature maps of the image into a spatially specific dynamic filtering operator, which performs spatially specific convolution operations on the image features to handle spatial-specific dominant degradations. By effectively integrating both global and local dynamic filtering operators, our proposed method outperforms state-of-the-art blind super-resolution algorithms in both synthetic and real image datasets.

Most super-resolution (SR) models struggle with real-world low-resolution (LR) images. This issue arises because the degradation characteristics in the synthetic datasets differ from those in real-world LR images. Since SR models are trained on pairs of high-resolution (HR) and LR images generated by downsampling, they are optimized for simple degradation. However, real-world LR images contain complex degradation caused by factors such as the imaging process and JPEG compression. Due to these differences in degradation characteristics, most SR models perform poorly on real-world LR images. This study proposes a dataset generation method using undertrained image reconstruction models. These models have the property of reconstructing low-quality images with diverse degradation from input images. By leveraging this property, this study generates LR images with diverse degradation from HR images to construct the datasets. Fine-tuning pre-trained SR models on our generated datasets improves noise removal and blur reduction, enhancing performance on real-world LR images. Furthermore, an analysis of the datasets reveals that degradation diversity contributes to performance improvements, whereas color differences between HR and LR images may degrade performance. 11 pages, (11 figures and 2 tables)

Existing Blind image Super-Resolution (BSR) methods focus on estimating either kernel or degradation information, but have long overlooked the essential content details. In this paper, we propose a novel BSR approach, Content-aware Degradation-driven Transformer (CDFormer), to capture both degradation and content representations. However, low-resolution images cannot provide enough content details, and thus we introduce a diffusion-based module $CDFormer_{diff}$ to first learn Content Degradation Prior (CDP) in both low- and high-resolution images, and then approximate the real distribution given only low-resolution information. Moreover, we apply an adaptive SR network $CDFormer_{SR}$ that effectively utilizes CDP to refine features. Compared to previous diffusion-based SR methods, we treat the diffusion model as an estimator that can overcome the limitations of expensive sampling time and excessive diversity. Experiments show that CDFormer can outperform existing methods, establishing a new state-of-the-art performance on various benchmarks under blind settings. Codes and models will be available at \href{https://github.com/I2-Multimedia-Lab/CDFormer}{https://github.com/I2-Multimedia-Lab/CDFormer}.

Self-supervised learning is crucial for super-resolution because ground-truth images are usually unavailable for real-world settings. Existing methods derive self-supervision from low-resolution images by creating pseudo-pairs or by enforcing a low-resolution reconstruction objective. These methods struggle with insufficient modeling of real-world degradations and the lack of knowledge about high-resolution imagery, resulting in unnatural super-resolved results. This paper strengthens awareness of the high-resolution image to improve the self-supervised real-world super-resolution. We propose a controller to adjust the degradation modeling based on the quality of super-resolution results. We also introduce a novel feature-alignment regularizer that directly constrains the distribution of super-resolved images. Our method finetunes the off-the-shelf SR models for a target real-world domain. Experiments show that it produces natural super-resolved images with state-of-the-art perceptual performance.

Recent Blind Image Super-Resolution (BSR) methods have shown proficiency in general images. However, we find that the efficacy of recent methods obviously diminishes when employed on image data with blur, while image data with intentional blur constitute a substantial proportion of general data. To further investigate and address this issue, we developed a new super-resolution dataset specifically tailored for blur images, named the Real-world Blur-kept Super-Resolution (ReBlurSR) dataset, which consists of nearly 3000 defocus and motion blur image samples with diverse blur sizes and varying blur intensities. Furthermore, we propose a new BSR framework for blur images called Perceptual-Blur-adaptive Super-Resolution (PBaSR), which comprises two main modules: the Cross Disentanglement Module (CDM) and the Cross Fusion Module (CFM). The CDM utilizes a dual-branch parallelism to isolate conflicting blur and general data during optimization. The CFM fuses the well-optimized prior from these distinct domains cost-effectively and efficiently based on model interpolation. By integrating these two modules, PBaSR achieves commendable performance on both general and blur data without any additional inference and deployment cost and is generalizable across multiple model architectures. Rich experiments show that PBaSR achieves state-of-the-art performance across various metrics without incurring extra inference costs. Within the widely adopted LPIPS metrics, PBaSR achieves an improvement range of approximately 0.02-0.10 with diverse anchor methods and blur types, across both the ReBlurSR and multiple common general BSR benchmarks. Code here: https://github.com/Imalne/PBaSR.

Super-resolution (SR) of satellite imagery is challenging due to the lack of paired low-/high-resolution data. Recent self-supervised SR methods overcome this limitation by exploiting the temporal redundancy in burst observations, but they lack a mechanism to quantify uncertainty in the reconstruction. In this work, we introduce a novel self-supervised loss that allows to estimate uncertainty in image super-resolution without ever accessing the ground-truth high-resolution data. We adopt a decision-theoretic perspective and show that minimizing the corresponding Bayesian risk yields the posterior mean and variance as optimal estimators. We validate our approach on a synthetic SkySat L1B dataset and demonstrate that it produces calibrated uncertainty estimates comparable to supervised methods. Our work bridges self-supervised restoration with uncertainty quantification, making a practical framework for uncertainty-aware image reconstruction.

Since non-blind Super Resolution (SR) fails to super-resolve Low-Resolution (LR) images degraded by arbitrary degradations, SR with the degradation model is required. However, this paper reveals that non-blind SR that is trained simply with various blur kernels exhibits comparable performance as those with the degradation model for blind SR. This result motivates us to revisit high-performance non-blind SR and extend it to blind SR with blur kernels. This paper proposes two SR networks by integrating kernel estimation and SR branches in an iterative end-to-end manner. In the first model, which is called the Kernel Conditioned Back-Projection Network (KCBPN), the low-dimensional kernel representations are estimated for conditioning the SR branch. In our second model, the Kernelized BackProjection Network (KBPN), a raw kernel is estimated and directly employed for modeling the image degradation. The estimated kernel is employed not only for back-propagating its residual but also for forward-propagating the residual to iterative stages. This forward-propagation encourages these stages to learn a variety of different features in different stages by focusing on pixels with large residuals in each stage. Experimental results validate the effectiveness of our proposed networks for kernel estimation and SR. We will release the code for this work.

Diffusion models (DM) have achieved remarkable promise in image super-resolution (SR). However, most of them are tailored to solving non-blind inverse problems with fixed known degradation settings, limiting their adaptability to real-world applications that involve complex unknown degradations. In this work, we propose BlindDiff, a DM-based blind SR method to tackle the blind degradation settings in SISR. BlindDiff seamlessly integrates the MAP-based optimization into DMs, which constructs a joint distribution of the low-resolution (LR) observation, high-resolution (HR) data, and degradation kernels for the data and kernel priors, and solves the blind SR problem by unfolding MAP approach along with the reverse process. Unlike most DMs, BlindDiff firstly presents a modulated conditional transformer (MCFormer) that is pre-trained with noise and kernel constraints, further serving as a posterior sampler to provide both priors simultaneously. Then, we plug a simple yet effective kernel-aware gradient term between adjacent sampling iterations that guides the diffusion model to learn degradation consistency knowledge. This also enables to joint refine the degradation model as well as HR images by observing the previous denoised sample. With the MAP-based reverse diffusion process, we show that BlindDiff advocates alternate optimization for blur kernel estimation and HR image restoration in a mutual reinforcing manner. Experiments on both synthetic and real-world datasets show that BlindDiff achieves the state-of-the-art performance with significant model complexity reduction compared to recent DM-based methods. Code will be available at \url{https://github.com/lifengcs/BlindDiff}

The problem of blind image super-resolution aims to recover high-resolution (HR) images from low-resolution (LR) images with unknown degradation modes. Most existing methods model the image degradation process using blur kernels. However, this explicit modeling approach struggles to cover the complex and varied degradation processes encountered in the real world, such as high-order combinations of JPEG compression, blur, and noise. Implicit modeling for the degradation process can effectively overcome this issue, but a key challenge of implicit modeling is the lack of accurate ground truth labels for the degradation process to conduct supervised training. To overcome this limitations inherent in implicit modeling, we propose an \textbf{U}ncertainty-based degradation representation for blind \textbf{S}uper-\textbf{R}esolution framework (\textbf{USR}). By suppressing the uncertainty of local degradation representations in images, USR facilitated self-supervised learning of degradation representations. The USR consists of two components: Adaptive Uncertainty-Aware Degradation Extraction (AUDE) and a feature extraction network composed of Variable Depth Dynamic Convolution (VDDC) blocks. To extract Uncertainty-based Degradation Representation from LR images, the AUDE utilizes the Self-supervised Uncertainty Contrast module with Uncertainty Suppression Loss to suppress the inherent model uncertainty of the Degradation Extractor. Furthermore, VDDC block integrates degradation information through dynamic convolution. Rhe VDDC also employs an Adaptive Intensity Scaling operation that adaptively adjusts the degradation representation according to the network hierarchy, thereby facilitating the effective integration of degradation information. Quantitative and qualitative experiments affirm the superiority of our approach.

The performance of image super-resolution relies heavily on the accuracy of degradation information, especially under blind settings. Due to the absence of true degradation models in real-world scenarios, previous methods learn distinct representations by distinguishing different degradations in a batch. However, the most significant degradation differences may provide shortcuts for the learning of representations such that subtle difference may be discarded. In this paper, we propose an alternative to learn degradation representations through reproducing degraded low-resolution (LR) images. By guiding the degrader to reconstruct input LR images, full degradation information can be encoded into the representations. In addition, we develop a distribution alignment loss to facilitate the learning of the degradation representations by introducing a bounded constraint. Moreover, to achieve larger receptive fields to capture information from a wider region for better restoration results, we introduce a degradation-aware Mamba module to efficiently model long-range dependency between the anchor pixel and the surrounding informative pixels. And the module strikes a flexible adaption to various degradations based on the learned representations. Experiments show that our representations can extract accurate and highly robust degradation information. Evaluations on both synthetic and real images demonstrate that our ReDSR achieves state-of-the-art performance for the blind SR tasks.

Recently, diffusion-based blind super-resolution (SR) methods have shown great ability to generate high-resolution images with abundant high-frequency detail, but the detail is often achieved at the expense of fidelity. Meanwhile, another line of research focusing on rectifying the reverse process of diffusion models (i.e., diffusion guidance), has demonstrated the power to generate high-fidelity results for non-blind SR. However, these methods rely on known degradation kernels, making them difficult to apply to blind SR. To address these issues, we present DADiff in this paper. DADiff incorporates degradation-aware models into the diffusion guidance framework, eliminating the need to know degradation kernels. Additionally, we propose two novel techniques: input perturbation and guidance scalar, to further improve our performance. Extensive experimental results show that our proposed method has superior performance over state-of-the-art methods on blind SR benchmarks.

Previous methods have demonstrated remarkable performance in single image super-resolution (SISR) tasks with known and fixed degradation (e.g., bicubic downsampling). However, when the actual degradation deviates from these assumptions, these methods may experience significant declines in performance. In this paper, we propose a Dual Branch Degradation Extractor Network to address the blind SR problem. While some blind SR methods assume noise-free degradation and others do not explicitly consider the presence of noise in the degradation model, our approach predicts two unsupervised degradation embeddings that represent blurry and noisy information. The SR network can then be adapted to blur embedding and noise embedding in distinct ways. Furthermore, we treat the degradation extractor as a regularizer to capitalize on differences between SR and HR images. Extensive experiments on several benchmarks demonstrate our method achieves SOTA performance in the blind SR problem.

Implicit degradation estimation-based blind super-resolution (IDE-BSR) hinges on extracting the implicit degradation representation (IDR) of the LR image and adapting it to LR image features to guide HR detail restoration. Although IDE-BSR has shown potential in dealing with noise interference and complex degradations, existing methods ignore the importance of IDR discriminability for BSR and instead over-complicate the adaptation process to improve effect, resulting in a significant increase in the model's parameters and computations. In this paper, we focus on the discriminability optimization of IDR and propose a new powerful and lightweight BSR model termed LightBSR. Specifically, we employ a knowledge distillation-based learning framework. We first introduce a well-designed degradation-prior-constrained contrastive learning technique during teacher stage to make the model more focused on distinguishing different degradation types. Then we utilize a feature alignment technique to transfer the degradation-related knowledge acquired by the teacher to the student for practical inferencing. Extensive experiments demonstrate the effectiveness of IDR discriminability-driven BSR model design. The proposed LightBSR can achieve outstanding performance with minimal complexity across a range of blind SR tasks. Our code is accessible at: https://github.com/MJ-NCEPU/LightBSR.