A unified learning framework for single image super-resolution

Heat Transfer-Inspired Network for Image Super-Resolution Reconstruction

Image super-resolution (SR) is a critical image preprocessing task for many applications. How to recover features as accurately as possible is the focus of SR algorithms. Most existing SR methods tend to guide the image reconstruction process with gradient maps, frequency perception modules, etc. and improve the quality of recovered images from the perspective of enhancing edges, but rarely optimize the neural network structure from the system level. In this article, we conduct an in- depth exploration for the inner nature of the SR network structure. In light of the consistency between thermal particles in the thermal field and pixels in the image domain, we propose a novel heat-transfer-inspired network (HTI-Net) for image SR reconstruction based on the theoretical basis of heat transfer. With the finite difference theory, we use a second-order mixed-difference equation to redesign the residual network (ResNet), which can fully integrate multiple information to achieve better feature reuse. In addition, according to the thermal conduction differential equation (TCDE) in the thermal field, the pixel value flow equation (PVFE) in the image domain is derived to mine deep potential feature information. The experimental results on multiple standard databases demonstrate that the proposed HTI-Net has superior edge detail reconstruction effect and parameter performance compared with the existing SR methods. The experimental results on the microscope chip image (MCI) database consisting of realistic low-resolution (LR) and high-resolution (HR) images show that the proposed HTI-Net for image SR reconstruction can improve the effectiveness of the hardware Trojan detection system.

Multi-scale Xception based depthwise separable convolution for single image super-resolution

The main target of Single image super-resolution is to recover high-quality or high-resolution image from degraded version of low-quality or low-resolution image. Recently, deep learning-based approaches have achieved significant performance in image super-resolution tasks. However, existing approaches related with image super-resolution fail to use the features information of low-resolution images as well as do not recover the hierarchical features for the final reconstruction purpose. In this research work, we have proposed a new architecture inspired by ResNet and Xception networks, which enable a significant drop in the number of network parameters and improve the processing speed to obtain the SR results. We are compared our proposed algorithm with existing state-of-the-art algorithms and confirmed the great ability to construct HR images with fine, rich, and sharp texture details as well as edges. The experimental results validate that our proposed approach has robust performance compared to other popular techniques related to accuracy, speed, and visual quality.

Learning Image Profile Enhancement and Denoising Statistics Priors for Single-Image Super-Resolution

Single-image super-resolution (SR) has been widely used in computer vision applications. The reconstruction-based SR methods are mainly based on certain prior terms to regularize the SR problem. However, it is very challenging to further improve the SR performance by the conventional design of explicit prior terms. Because of the powerful learning ability, deep convolutional neural networks (CNNs) have been widely used in single-image SR task. However, it is difficult to achieve further improvement by only designing the network architecture. In addition, most existing deep CNN-based SR methods learn a nonlinear mapping function to directly map low-resolution (LR) images to desirable high-resolution (HR) images, ignoring the observation models of input images. Inspired by the split Bregman iteration (SBI) algorithm, which is a powerful technique for solving the constrained optimization problems, the original SR problem is divided into two subproblems: 1) inversion subproblem and 2) denoising subproblem. Since the inversion subproblem can be regarded as an inversion step to reconstruct an intermediate HR image with sharper edges and finer structures, we propose to use deep CNN to capture low-level explicit image profile enhancement prior (PEP). Since the denoising subproblem aims to remove the noise in the intermediate image, we adopt a simple and effective denoising network to learn implicit image denoising statistics prior (DSP). Furthermore, the penalty parameter in SBI is adaptively tuned during the iterations for better performance. Finally, we also prove the convergence of our method. Thus, the deep CNNs are exploited to capture both implicit and explicit image statistics priors. Due to SBI, the SR observation model is also leveraged. Consequently, it bridges between two popular SR approaches: 1) learning-based method and 2) reconstruction-based method. Experimental results show that the proposed method achieves the state-of-the-art SR results.

Generating Face Images With Attributes for Free

With superhuman-level performance of face recognition, we are more concerned about the recognition of fine-grained attributes, such as emotion, age, and gender. However, given that the label space is extremely large and follows a long-tail distribution, it is quite expensive to collect sufficient samples for fine-grained attributes. This results in imbalanced training samples and inferior attribute recognition models. To this end, we propose the use of arbitrary attribute combinations, without human effort, to synthesize face images. In particular, to bridge the semantic gap between high-level attribute label space and low-level face image, we propose a novel neural-network-based approach that maps the target attribute labels to an embedding vector, which can be fed into a pretrained image decoder to synthesize a new face image. Furthermore, to regularize the attribute for image synthesis, we propose to use a perceptual loss to make the new image explicitly faithful to target attributes. Experimental results show that our approach can generate photorealistic face images from attribute labels, and more importantly, by serving as augmented training samples, these images can significantly boost the performance of attribute recognition model. The code is open-sourced at this link.

Face Mask Assistant: Detection of Face Mask Service Stage Based on Mobile Phone

Coronavirus Disease 2019 (COVID-19) has spread all over the world since it broke out massively in December 2019, which has caused a large loss to the whole world. Both the confirmed cases and death cases have reached a relatively frightening number. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the cause of COVID-19, can be transmitted by small respiratory droplets. To curb its spread at the source, wearing masks is a convenient and effective measure. In most cases, people use face masks in a high-frequent but short-time way. Aimed at solving the problem that we do not know which service stage of the mask belongs to, we propose a detection system based on the mobile phone. We first extract four features from the gray level co-occurrence matrixes (GLCMs) of the face mask's micro-photos. Next, a three-result detection system is accomplished by using K Nearest Neighbor (KNN) algorithm. The results of validation experiments show that our system can reach an accuracy of 82.87% (measured by macro-measures) on the testing dataset. The precision of Type I 'normal use' and the recall of type III 'not recommended' reach 92.00% and 92.59%. In future work, we plan to expand the detection objects to more mask types. This work demonstrates that the proposed mobile microscope system can be used as an assistant for face mask being used, which may play a positive role in fighting against COVID-19.

Survey on Multi-Output Learning

The aim of multi-output learning is to simultaneously predict multiple outputs given an input. It is an important learning problem for decision-making since making decisions in the real world often involves multiple complex factors and criteria. In recent times, an increasing number of research studies have focused on ways to predict multiple outputs at once. Such efforts have transpired in different forms according to the particular multi-output learning problem under study. Classic cases of multi-output learning include multi-label learning, multi-dimensional learning, multi-target regression, and others. From our survey of the topic, we were struck by a lack in studies that generalize the different forms of multi-output learning into a common framework. This article fills that gap with a comprehensive review and analysis of the multi-output learning paradigm. In particular, we characterize the four Vs of multi-output learning, i.e., volume, velocity, variety, and veracity, and the ways in which the four Vs both benefit and bring challenges to multi-output learning by taking inspiration from big data. We analyze the life cycle of output labeling, present the main mathematical definitions of multi-output learning, and examine the field's key challenges and corresponding solutions as found in the literature. Several model evaluation metrics and popular data repositories are also discussed. Last but not least, we highlight some emerging challenges with multi-output learning from the perspective of the four Vs as potential research directions worthy of further studies.

Enhanced Non-Local Total Variation Model and Multi-Directional Feature Prediction Prior for Single Image Super Resolution

It is widely acknowledged that single image super-resolution (SISR) methods play a critical role in recovering the missing high-frequencies in an input low-resolution image. As SISR is severely ill-conditioned, image priors are necessary to regularize the solution spaces and generate the corresponding high-resolution image. In this paper, we propose an effective SISR framework based on the enhanced non-local similarity modeling and learning-based multi-directional feature prediction (ENLTV-MDFP). Since both the modeled and learned priors are exploited, the proposed ENLTV-MDFP method benefits from the complementary properties of the reconstruction-based and learning-based SISR approaches. Specifically, for the non-local similarity-based modeled prior [enhanced non-local total variation, (ENLTV)], it is characterized via the decaying kernel and stable group similarity reliability schemes. For the learned prior [multi-directional feature prediction prior, (MDFP)], it is learned via the deep convolutional neural network. The modeled prior performs well in enhancing edges and suppressing visual artifacts, while the learned prior is effective in hallucinating details from external images. Combining these two complementary priors in the MAP framework, a combined SR cost function is proposed. Finally, the combined SR problem is solved via the split Bregman iteration algorithm. Based on the extensive experiments, the proposed ENLTV-MDFP method outperforms many state-of-the-art algorithms visually and quantitatively.

Robust Regression Estimation Based on Low-Dimensional Recurrent Neural Networks

The robust Huber's M-estimator is widely used in signal and image processing, classification, and regression. From an optimization point of view, Huber's M-estimation problem is often formulated as a large-sized quadratic programming (QP) problem in view of its nonsmooth cost function. This paper presents a generalized regression estimator which minimizes a reduced-sized QP problem. The generalized regression estimator may be viewed as a significant generalization of several robust regression estimators including Huber's M-estimator. The performance of the generalized regression estimator is analyzed in terms of robustness and approximation accuracy. Furthermore, two low-dimensional recurrent neural networks (RNNs) are introduced for robust estimation. The two RNNs have low model complexity and enhanced computational efficiency. Finally, the experimental results of two examples and an application to image restoration are presented to substantiate superior performance of the proposed method over conventional algorithms for robust regression estimation in terms of approximation accuracy and convergence rate.

Matrix function optimization under weighted boundary constraints and its applications in network control

The matrix function optimization under weighted boundary constraints on the matrix variables is investigated in this work. An "index-notation-arrangement based chain rule" (I-Chain rule) is introduced to obtain the gradient of a matrix function. By doing this, we propose the weighted trace-constraint-based projected gradient method (WTPGM) and weighted orthornormal-constraint-based projected gradient method (WOPGM) to locate a point of minimum of an objective/cost function of matrix variables iteratively subject to weighted trace constraint and weighted orthonormal constraint, respectively. New techniques are implemented to establish the convergence property of both algorithms. In addition, compared with the existing scheme termed "orthornormal-constraint-based projected gradient method" (OPGM) that requires the gradient has to be represented by the multiplication of a symmetrical matrix and the matrix variable itself, such a condition has been relaxed in WOPGM. Simulation results show the effectiveness of our methods not only in network control but also in other learning problems. We believe that the results reveal interesting physical insights in the field of network control and allow extensive applications of matrix function optimization problems in science and engineering.

Efficient kNN Classification With Different Numbers of Nearest Neighbors

nearest neighbor (kNN) method is a popular classification method in data mining and statistics because of its simple implementation and significant classification performance. However, it is impractical for traditional kNN methods to assign a fixed value (even though set by experts) to all test samples. Previous solutions assign different values to different test samples by the cross validation method but are usually time-consuming. This paper proposes a kTree method to learn different optimal values for different test/new samples, by involving a training stage in the kNN classification. Specifically, in the training stage, kTree method first learns optimal values for all training samples by a new sparse reconstruction model, and then constructs a decision tree (namely, kTree) using training samples and the learned optimal values. In the test stage, the kTree fast outputs the optimal value for each test sample, and then, the kNN classification can be conducted using the learned optimal value and all training samples. As a result, the proposed kTree method has a similar running cost but higher classification accuracy, compared with traditional kNN methods, which assign a fixed value to all test samples. Moreover, the proposed kTree method needs less running cost but achieves similar classification accuracy, compared with the newly kNN methods, which assign different values to different test samples. This paper further proposes an improvement version of kTree method (namely, k*Tree method) to speed its test stage by extra storing the information of the training samples in the leaf nodes of kTree, such as the training samples located in the leaf nodes, their kNNs, and the nearest neighbor of these kNNs. We call the resulting decision tree as k*Tree, which enables to conduct kNN classification using a subset of the training samples in the leaf nodes rather than all training samples used in the newly kNN methods. This actually reduces running cost of test stage. Finally, the experimental results on 20 real data sets showed that our proposed methods (i.e., kTree and k*Tree) are much more efficient than the compared methods in terms of classification tasks.

Robust Joint Graph Sparse Coding for Unsupervised Spectral Feature Selection

In this paper, we propose a new unsupervised spectral feature selection model by embedding a graph regularizer into the framework of joint sparse regression for preserving the local structures of data. To do this, we first extract the bases of training data by previous dictionary learning methods and, then, map original data into the basis space to generate their new representations, by proposing a novel joint graph sparse coding (JGSC) model. In JGSC, we first formulate its objective function by simultaneously taking subspace learning and joint sparse regression into account, then, design a new optimization solution to solve the resulting objective function, and further prove the convergence of the proposed solution. Furthermore, we extend JGSC to a robust JGSC (RJGSC) via replacing the least square loss function with a robust loss function, for achieving the same goals and also avoiding the impact of outliers. Finally, experimental results on real data sets showed that both JGSC and RJGSC outperformed the state-of-the-art algorithms in terms of k -nearest neighbor classification performance.

Coarse-to-Fine Learning for Single-Image Super-Resolution

This paper develops a coarse-to-fine framework for single-image super-resolution (SR) reconstruction. The coarse-to-fine approach achieves high-quality SR recovery based on the complementary properties of both example learning-and reconstruction-based algorithms: example learning-based SR approaches are useful for generating plausible details from external exemplars but poor at suppressing aliasing artifacts, while reconstruction-based SR methods are propitious for preserving sharp edges yet fail to generate fine details. In the coarse stage of the method, we use a set of simple yet effective mapping functions, learned via correlative neighbor regression of grouped low-resolution (LR) to high-resolution (HR) dictionary atoms, to synthesize an initial SR estimate with particularly low computational cost. In the fine stage, we devise an effective regularization term that seamlessly integrates the properties of local structural regularity, nonlocal self-similarity, and collaborative representation over relevant atoms in a learned HR dictionary, to further improve the visual quality of the initial SR estimation obtained in the coarse stage. The experimental results indicate that our method outperforms other state-of-the-art methods for producing high-quality images despite that both the initial SR estimation and the followed enhancement are cheap to implement.

Pairwise Operator Learning for Patch-Based Single-Image Super-Resolution

Motivated by the fact that image patches could be inherently represented by matrices, single-image super-resolution is treated as a problem of learning regression operators in a matrix space in this paper. The regression operators that map low-resolution image patches to high-resolution image patches are generally defined by the left and right multiplication operators. The pairwise operators are, respectively, used to extract the raw and column information of low-resolution image patches for recovering high-resolution estimations. The patch-based regression algorithm possesses three favorable properties. First, the proposed super-resolution algorithm is efficient during both training and testing, because image patches are treated as matrices. Second, the data storage requirement of the optimal pairwise operator is far less than most popular single-image super-resolution algorithms, because only two small sized matrices need to be stored. Last, the super-resolution performance is competitive with most popular single-image super-resolution algorithms, because both raw and column information of image patches is considered. Experimental results show the efficiency and effectiveness of the proposed patch-based single-image super-resolution algorithm.

Coupled Deep Autoencoder for Single Image Super-Resolution

Sparse coding has been widely applied to learning-based single image super-resolution (SR) and has obtained promising performance by jointly learning effective representations for low-resolution (LR) and high-resolution (HR) image patch pairs. However, the resulting HR images often suffer from ringing, jaggy, and blurring artifacts due to the strong yet ad hoc assumptions that the LR image patch representation is equal to, is linear with, lies on a manifold similar to, or has the same support set as the corresponding HR image patch representation. Motivated by the success of deep learning, we develop a data-driven model coupled deep autoencoder (CDA) for single image SR. CDA is based on a new deep architecture and has high representational capability. CDA simultaneously learns the intrinsic representations of LR and HR image patches and a big-data-driven function that precisely maps these LR representations to their corresponding HR representations. Extensive experimentation demonstrates the superior effectiveness and efficiency of CDA for single image SR compared to other state-of-the-art methods on Set5 and Set14 datasets.

Similarity Constraints-Based Structured Output Regression Machine: An Approach to Image Super-Resolution

For regression-based single-image super-resolution (SR) problem, the key is to establish a mapping relation between high-resolution (HR) and low-resolution (LR) image patches for obtaining a visually pleasing quality image. Most existing approaches typically solve it by dividing the model into several single-output regression problems, which obviously ignores the circumstance that a pixel within an HR patch affects other spatially adjacent pixels during the training process, and thus tends to generate serious ringing artifacts in resultant HR image as well as increase computational burden. To alleviate these problems, we propose to use structured output regression machine (SORM) to simultaneously model the inherent spatial relations between the HR and LR patches, which is propitious to preserve sharp edges. In addition, to further improve the quality of reconstructed HR images, a nonlocal (NL) self-similarity prior in natural images is introduced to formulate as a regularization term to further enhance the SORM-based SR results. To offer a computation-effective SORM method, we use a relative small nonsupport vector samples to establish the accurate regression model and an accelerating algorithm for NL self-similarity calculation. Extensive SR experiments on various images indicate that the proposed method can achieve more promising performance than the other state-of-the-art SR methods in terms of both visual quality and computational cost.

SERF: A Simple, Effective, Robust, and Fast Image Super-Resolver From Cascaded Linear Regression

Example learning-based image super-resolution techniques estimate a high-resolution image from a low-resolution input image by relying on high- and low-resolution image pairs. An important issue for these techniques is how to model the relationship between high- and low-resolution image patches: most existing complex models either generalize hard to diverse natural images or require a lot of time for model training, while simple models have limited representation capability. In this paper, we propose a simple, effective, robust, and fast (SERF) image super-resolver for image super-resolution. The proposed super-resolver is based on a series of linear least squares functions, namely, cascaded linear regression. It has few parameters to control the model and is thus able to robustly adapt to different image data sets and experimental settings. The linear least square functions lead to closed form solutions and therefore achieve computationally efficient implementations. To effectively decrease these gaps, we group image patches into clusters via k-means algorithm and learn a linear regressor for each cluster at each iteration. The cascaded learning process gradually decreases the gap of high-frequency detail between the estimated high-resolution image patch and the ground truth image patch and simultaneously obtains the linear regression parameters. Experimental results show that the proposed method achieves superior performance with lower time consumption than the state-of-the-art methods.

An Example-Based Super-Resolution Algorithm for Selfie Images

A selfie is typically a self-portrait captured using the front camera of a smartphone. Most state-of-the-art smartphones are equipped with a high-resolution (HR) rear camera and a low-resolution (LR) front camera. As selfies are captured by front camera with limited pixel resolution, the fine details in it are explicitly missed. This paper aims to improve the resolution of selfies by exploiting the fine details in HR images captured by rear camera using an example-based super-resolution (SR) algorithm. HR images captured by rear camera carry significant fine details and are used as an exemplar to train an optimal matrix-value regression (MVR) operator. The MVR operator serves as an image-pair priori which learns the correspondence between the LR-HR patch-pairs and is effectively used to super-resolve LR selfie images. The proposed MVR algorithm avoids vectorization of image patch-pairs and preserves image-level information during both learning and recovering process. The proposed algorithm is evaluated for its efficiency and effectiveness both qualitatively and quantitatively with other state-of-the-art SR algorithms. The results validate that the proposed algorithm is efficient as it requires less than 3 seconds to super-resolve LR selfie and is effective as it preserves sharp details without introducing any counterfeit fine details.

Single Image Superresolution via Directional Group Sparsity and Directional Features

Single image superresolution (SR) aims to construct a high-resolution version from a single low-resolution (LR) image. The SR reconstruction is challenging because of the missing details in the given LR image. Thus, it is critical to explore and exploit effective prior knowledge for boosting the reconstruction performance. In this paper, we propose a novel SR method by exploiting both the directional group sparsity of the image gradients and the directional features in similarity weight estimation. The proposed SR approach is based on two observations: 1) most of the sharp edges are oriented in a limited number of directions and 2) an image pixel can be estimated by the weighted averaging of its neighbors. In consideration of these observations, we apply the curvelet transform to extract directional features which are then used for region selection and weight estimation. A combined total variation regularizer is presented which assumes that the gradients in natural images have a straightforward group sparsity structure. In addition, a directional nonlocal means regularization term takes pixel values and directional information into account to suppress unwanted artifacts. By assembling the designed regularization terms, we solve the SR problem of an energy function with minimal reconstruction error by applying a framework of templates for first-order conic solvers. The thorough quantitative and qualitative results in terms of peak signal-to-noise ratio, structural similarity, information fidelity criterion, and preference matrix demonstrate that the proposed approach achieves higher quality SR reconstruction than the state-of-the-art algorithms.

Blind image quality assessment via deep learning

This paper investigates how to blindly evaluate the visual quality of an image by learning rules from linguistic descriptions. Extensive psychological evidence shows that humans prefer to conduct evaluations qualitatively rather than numerically. The qualitative evaluations are then converted into the numerical scores to fairly benchmark objective image quality assessment (IQA) metrics. Recently, lots of learning-based IQA models are proposed by analyzing the mapping from the images to numerical ratings. However, the learnt mapping can hardly be accurate enough because some information has been lost in such an irreversible conversion from the linguistic descriptions to numerical scores. In this paper, we propose a blind IQA model, which learns qualitative evaluations directly and outputs numerical scores for general utilization and fair comparison. Images are represented by natural scene statistics features. A discriminative deep model is trained to classify the features into five grades, corresponding to five explicit mental concepts, i.e., excellent, good, fair, poor, and bad. A newly designed quality pooling is then applied to convert the qualitative labels into scores. The classification framework is not only much more natural than the regression-based models, but also robust to the small sample size problem. Thorough experiments are conducted on popular databases to verify the model's effectiveness, efficiency, and robustness.

Learning multiple linear mappings for efficient single image super-resolution

Example learning-based superresolution (SR) algorithms show promise for restoring a high-resolution (HR) image from a single low-resolution (LR) input. The most popular approaches, however, are either time- or space-intensive, which limits their practical applications in many resource-limited settings. In this paper, we propose a novel computationally efficient single image SR method that learns multiple linear mappings (MLM) to directly transform LR feature subspaces into HR subspaces. In particular, we first partition the large nonlinear feature space of LR images into a cluster of linear subspaces. Multiple LR subdictionaries are then learned, followed by inferring the corresponding HR subdictionaries based on the assumption that the LR-HR features share the same representation coefficients. We establish MLM from the input LR features to the desired HR outputs in order to achieve fast yet stable SR recovery. Furthermore, in order to suppress displeasing artifacts generated by the MLM-based method, we apply a fast nonlocal means algorithm to construct a simple yet effective similarity-based regularization term for SR enhancement. Experimental results indicate that our approach is both quantitatively and qualitatively superior to other application-oriented SR methods, while maintaining relatively low time and space complexity.