Showing papers by "Harbin Institute of Technology published in 2015"

Document Modeling with Gated Recurrent Neural Network for Sentiment Classification

Abstract: Document level sentiment classification remains a challenge: encoding the intrinsic relations between sentences in the semantic meaning of a document. To address this, we introduce a neural network model to learn vector-based document representation in a unified, bottom-up fashion. The model first learns sentence representation with convolutional neural network or long short-term memory. Afterwards, semantics of sentences and their relations are adaptively encoded in document representation with gated recurrent neural network. We conduct document level sentiment classification on four large-scale review datasets from IMDB and Yelp Dataset Challenge. Experimental results show that: (1) our neural model shows superior performances over several state-of-the-art algorithms; (2) gated recurrent neural network dramatically outperforms standard recurrent neural network in document modeling for sentiment classification. 1

Spectral–Spatial Classification of Hyperspectral Data Based on Deep Belief Network

Abstract: Hyperspectral data classification is a hot topic in remote sensing community. In recent years, significant effort has been focused on this issue. However, most of the methods extract the features of original data in a shallow manner. In this paper, we introduce a deep learning approach into hyperspectral image classification. A new feature extraction (FE) and image classification framework are proposed for hyperspectral data analysis based on deep belief network (DBN). First, we verify the eligibility of restricted Boltzmann machine (RBM) and DBN by the following spectral information-based classification. Then, we propose a novel deep architecture, which combines the spectral–spatial FE and classification together to get high classification accuracy. The framework is a hybrid of principal component analysis (PCA), hierarchical learning-based FE, and logistic regression (LR). Experimental results with hyperspectral data indicate that the classifier provide competitive solution with the state-of-the-art methods. In addition, this paper reveals that deep learning system has huge potential for hyperspectral data classification.

A Survey of Sparse Representation: Algorithms and Applications

Abstract: Sparse representation has attracted much attention from researchers in fields of signal processing, image processing, computer vision, and pattern recognition. Sparse representation also has a good reputation in both theoretical research and practical applications. Many different algorithms have been proposed for sparse representation. The main purpose of this paper is to provide a comprehensive study and an updated review on sparse representation and to supply guidance for researchers. The taxonomy of sparse representation methods can be studied from various viewpoints. For example, in terms of different norm minimizations used in sparsity constraints, the methods can be roughly categorized into five groups: 1) sparse representation with $l_{0}$ -norm minimization; 2) sparse representation with $l_{p}$ -norm ( $0 ) minimization; 3) sparse representation with $l_{1}$ -norm minimization; 4) sparse representation with $l_{2,1}$ -norm minimization; and 5) sparse representation with $l_{2}$ -norm minimization. In this paper, a comprehensive overview of sparse representation is provided. The available sparse representation algorithms can also be empirically categorized into four groups: 1) greedy strategy approximation; 2) constrained optimization; 3) proximity algorithm-based optimization; and 4) homotopy algorithm-based sparse representation. The rationales of different algorithms in each category are analyzed and a wide range of sparse representation applications are summarized, which could sufficiently reveal the potential nature of the sparse representation theory. In particular, an experimentally comparative study of these sparse representation algorithms was presented.

Activation of Peroxymonosulfate by Benzoquinone: A Novel Nonradical Oxidation Process

Abstract: The reactions between peroxymonosulfate (PMS) and quinones were investigated for the first time in this work, where benzoquinone (BQ) was selected as a model quinone. It was demonstrated that BQ could efficiently activate PMS for the degradation of sulfamethoxazole (SMX; a frequently detected antibiotic in the environments), and the degradation rate increased with solution pH from 7 to 10. Interestingly, quenching studies suggested that neither hydroxyl radical (•OH) nor sulfate radical (SO4•–) was produced therein. Instead, the generation of singlet oxygen (1O2) was proved by using two chemical probes (i.e., 2,2,6,6-tetramethyl-4-piperidinol and 9,10-diphenylanthracene) with the appearance of 1O2 indicative products detected by electron paramagnetic resonance spectrometry and liquid chromatography mass spectrometry, respectively. A catalytic mechanism was proposed involving the formation of a dioxirane intermediate between PMS and BQ and the subsequent decomposition of this intermediate into 1O2. Accordi...

Data-Based Techniques Focused on Modern Industry: An Overview

Abstract: This paper provides an overview of the recent developments in data-based techniques focused on modern industrial applications. As one of the hottest research topics for complicated processes, the data-based techniques have been rapidly developed over the past two decades and widely used in numerous industrial sectors nowadays. The core of data-based techniques is to take full advantage of the huge amounts of available process data, aiming to acquire the useful information within. Compared with the well-developed model-based approaches, data-based techniques provide efficient alternative solutions for different industrial issues under various operating conditions. The main objective of this paper is to review and summarize the recent achievements in data-based techniques, especially for complicated industrial applications, thus providing a referee for further study on the related topics both from academic and practical points of view. This paper begins with a brief evolutionary overview of data-based techniques in the last two decades. Then, the methodologies only based on process measurements and the model-data integrated techniques will be further introduced. The recent developments for modern industrial applications are, respectively, presented mainly from perspectives of monitoring and control. The new trends of data-based technique as well as potential application fields are finally discussed.

Sulfate radicals induced from peroxymonosulfate by magnetic ferrospinel MFe2O4 (M = Co, Cu, Mn, and Zn) as heterogeneous catalysts in the water

Abstract: Magnetic ferrospinel MFe 2 O 4 (M = Co, Cu, Mn, and Zn) prepared in a sol–gel process was introduced as catalyst to generate powerful radicals from peroxymonosulfate (PMS) for refractory di-n-butyl phthalate (DBP) degradation in the water. Various catalysts were described and characterized, and the catalytic activities in PMS oxidation system were investigated. Most important of all, the mechanism of different catalysts in the catalytic PMS solution was illustrated. The results showed that the incorporation of CoFe 2 O 4 had the highest catalytic performance in PMS oxidation for DBP degradation. All catalysts presented favorable recycling and stability in the repeated batch experiment. The catalytic process showed a dependence on initial pH, and an uncharged surface of the catalyst was more profitable for sulfate radical generation. H 2 -TPR and CVs analysis indicated that the sequence of the catalyst's reducibility in PMS solution was CoFe 2 O 4 > CuFe 2 O 4 > MnFe 2 O 4 > ZnFe 2 O 4 , which had a close connection with the activity of metal ion in A site of the catalysts. The surface hydroxyl sites played an important role in the catalytic process, and its quantity determined the degradation of DBP. Moreover, the reactive species in PMS/MFe 2 O 4 system were identified as sulfate radical and hydroxyl radical. The promotion of these radical's reaction was due to the fact that a balance action in the process of M 2+ /M 3+ , O 2− /O 2 , occurred, and at the same time, PMS was catalyzed.

Pumped hydro energy storage system: A technological review

Abstract: The pumped hydro energy storage (PHES) is a well-established and commercially-acceptable technology for utility-scale electricity storage and has been used since as early as the 1890s. Hydro power is not only a renewable and sustainable energy source, but its flexibility and storage capacity also make it possible to improve grid stability and to support the deployment of other intermittent renewable energy sources such as wind and solar. As a result, a renewed interest in PHES and a demand for the rehabilitation of old small hydro power plants are emerging globally. With regard to PHES, advances in turbine design are required to enhance plant performance and flexibility and new strategies for optimizing storage capacity and for maximizing plant profitability in the deregulated energy market. In the early 2000s, this technology has again emerged as an economically and technologically acceptable option for peak load shaving and wind and solar energy storage for power quality assurance. Furthermore, renewable energy sources due to their fluctuating nature cannot maintain or regulate continuous supply of power and hence require bulk electricity storage. The present study aims at reviewing the existing global PHES capacities, technological development, and hybrid systems (wind-hydro, solar pv-hydro, and wind-pv-hydro) and recommending the best possible options. The review explores that PHES is the most suitable technology for small autonomous island grids and massive energy storage, where the energy efficiency of PHES varies in practice between 70% and 80% with some claiming up to 87%. Around the world, PHES size mostly nestles in the range of 1000–1500 MW, being as large as 2000–3000 MW. On the other hand, photovoltaic based pumped storage systems have been used for very small scale (load of few houses) only.

Neutrino Physics with JUNO

Abstract: The Jiangmen Underground Neutrino Observatory (JUNO), a 20 kton multi-purpose underground liquid scintillator detector, was proposed with the determination of the neutrino mass hierarchy as a primary physics goal. It is also capable of observing neutrinos from terrestrial and extra-terrestrial sources, including supernova burst neutrinos, diffuse supernova neutrino background, geoneutrinos, atmospheric neutrinos, solar neutrinos, as well as exotic searches such as nucleon decays, dark matter, sterile neutrinos, etc. We present the physics motivations and the anticipated performance of the JUNO detector for various proposed measurements. By detecting reactor antineutrinos from two power plants at 53-km distance, JUNO will determine the neutrino mass hierarchy at a 3-4 sigma significance with six years of running. The measurement of antineutrino spectrum will also lead to the precise determination of three out of the six oscillation parameters to an accuracy of better than 1\%. Neutrino burst from a typical core-collapse supernova at 10 kpc would lead to ~5000 inverse-beta-decay events and ~2000 all-flavor neutrino-proton elastic scattering events in JUNO. Detection of DSNB would provide valuable information on the cosmic star-formation rate and the average core-collapsed neutrino energy spectrum. Geo-neutrinos can be detected in JUNO with a rate of ~400 events per year, significantly improving the statistics of existing geoneutrino samples. The JUNO detector is sensitive to several exotic searches, e.g. proton decay via the $p\to K^++\bar u$ decay channel. The JUNO detector will provide a unique facility to address many outstanding crucial questions in particle and astrophysics. It holds the great potential for further advancing our quest to understanding the fundamental properties of neutrinos, one of the building blocks of our Universe.

Lightweight, Superelastic, and Mechanically Flexible Graphene/Polyimide Nanocomposite Foam for Strain Sensor Application.

Abstract: The creation of superelastic, flexible three-dimensional (3D) graphene-based architectures is still a great challenge due to structure collapse or significant plastic deformation. Herein, we report a facile approach of transforming the mechanically fragile reduced graphene oxide (rGO) aerogel into superflexible 3D architectures by introducing water-soluble polyimide (PI). The rGO/PI nanocomposites are fabricated using strategies of freeze casting and thermal annealing. The resulting monoliths exhibit low density, excellent flexibility, superelasticity with high recovery rate, and extraordinary reversible compressibility. The synergistic effect between rGO and PI endows the elastomer with desirable electrical conductivity, remarkable compression sensitivity, and excellent durable stability. The rGO/PI nanocomposites show potential applications in multifunctional strain sensors under the deformations of compression, bending, stretching, and torsion.

Deep learning for event-driven stock prediction

Abstract: We propose a deep learning method for event-driven stock market prediction. First, events are extracted from news text, and represented as dense vectors, trained using a novel neural tensor network. Second, a deep convolutional neural network is used to model both short-term and long-term influences of events on stock price movements. Experimental results show that our model can achieve nearly 6% improvements on S&P 500 index prediction and individual stock prediction, respectively, compared to state-of-the-art baseline methods. In addition, market simulation results show that our system is more capable of making profits than previously reported systems trained on S&P 500 stock historical data.

Hierarchical β‐Mo2C Nanotubes Organized by Ultrathin Nanosheets as a Highly Efficient Electrocatalyst for Hydrogen Production

Abstract: Production of hydrogen by electrochemical water splitting has been hindered by the high cost of precious metal catalysts, such as Pt, for the hydrogen evolution reaction (HER). In this work, novel hierarchical β-Mo2C nanotubes constructed from porous nanosheets have been fabricated and investigated as a high-performance and low-cost electrocatalyst for HER. An unusual template-engaged strategy has been utilized to controllably synthesize Mo-polydopamine nanotubes, which are further converted into hierarchical β-Mo2C nanotubes by direct carburization at high temperature. Benefitting from several structural advantages including ultrafine primary nanocrystallites, large exposed surface, fast charge transfer, and unique tubular structure, the as-prepared hierarchical β-Mo2C nanotubes exhibit excellent electrocatalytic performance for HER with small overpotential in both acidic and basic conditions, as well as remarkable stability.

Metal organic framework-derived Fe/C nanocubes toward efficient microwave absorption

Abstract: Composites of magnetic metal nanoparticles and carbon materials are highly desirable for high-performance microwave absorbers due to their compatible dielectric loss and magnetic loss abilities. In this article, novel nanocomposites, Fe/C nanocubes, have been successfully prepared through an in situ route from a metal–organic framework, Prussian blue, by controlled high-temperature pyrolysis. The resultant nanocubes are actually composed of a cubic framework of amorphous carbon and uniformly dispersed core–shell Fe@graphitic carbon nanoparticles. Within the studied pyrolysis temperature range (600–700 °C), the porous structure, iron content, magnetic properties, and graphitization degree of the Fe/C nanocubes can be well modulated. Particularly, the improved carbon graphitization degree, both in amorphous frameworks and graphitic shells, results in enhanced complex permittivity and dielectric loss properties. The homogeneous chemical composition and microstructure stimulate the formation of multiple dielectric resonances by regularizing various polarizations. The synergistic effect of dielectric loss, magnetic loss, matched impedance, and dielectric resonances accounts for the improved microwave absorption properties of the Fe/C nanocubes. The absorption bands of the optimum one obtained at 650 °C are superior to most composites ever reported. By considering the good chemical homogeneity and microwave absorption, we believe that the as-fabricated Fe/C nanocubes will be promising candidates as highly effective microwave absorbers.

Precision Measurement of the Helium Flux in Primary Cosmic Rays of Rigidities 1.9 GV to 3 TV with the Alpha Magnetic Spectrometer on the International Space Station

Abstract: Knowledge of the precise rigidity dependence of the helium flux is important in understanding the origin, acceleration, and propagation of cosmic rays. A precise measurement of the helium flux in primary cosmic rays with rigidity (momentum/charge) from 1.9 GV to 3 TV based on 50 million events is presented and compared to the proton flux. The detailed variation with rigidity of the helium flux spectral index is presented for the first time. The spectral index progressively hardens at rigidities larger than 100 GV. The rigidity dependence of the helium flux spectral index is similar to that of the proton spectral index though the magnitudes are different. Remarkably, the spectral index of the proton to helium flux ratio increases with rigidity up to 45 GV and then becomes constant; the flux ratio above 45 GV is well described by a single power law.

Light upconverting core–shell nanostructures: nanophotonic control for emerging applications

Abstract: Light upconverting nanostructures employing lanthanide ions constitute an emerging research field recognized with wide ramifications and impact in many areas ranging from healthcare, to energy and, to security. The core–shell design of these nanostructures allows us to deliberately introduce a hierarchy of electronic energy states, thus providing unprecedented opportunities to manipulate the electronic excitation, energy transfer and upconverted emissions. The core–shell morphology also causes the suppression of quenching mechanisms to produce efficient upconversion emission for biophotonic and photonic applications. Using hierarchical architect, whereby each shell layer can be defined to have a specific feature, the electronic structure as well as the physiochemical structure of the upconverting nanomaterials can be tuned to couple other electronic states on the surface such as excitations of organic dye molecules or localized surface plasmons from metallic nanostructures, or to introduce a broad range of imaging or therapeutic modalities into a single conduct. In this review, we summarize the key aspects of nanophotonic control of the light upconverting nanoparticles through governed design and preparation of hierarchical shells in the core–shell nanostructures, and review their emerging applications in the biomedical field, solar energy conversion, as well as security encoding.

Production of Sulfate Radical and Hydroxyl Radical by Reaction of Ozone with Peroxymonosulfate: A Novel Advanced Oxidation Process

Abstract: In this work, simultaneous generation of hydroxyl radical (•OH) and sulfate radical (SO4•–) by the reaction of ozone (O3) with peroxymonosulfate (PMS; HSO5–) has been proposed and experimentally verified. We demonstrate that the reaction between the anion of PMS (i.e., SO52–) and O3 is primarily responsible for driving O3 consumption with a measured second order rate constant of (2.12 ± 0.03) × 104 M–1 s–1. The formation of both •OH and SO4•– from the reaction between SO52– and O3 is confirmed by chemical probes (i.e., nitrobenzene for •OH and atrazine for both •OH and SO4•–). The yields of •OH and SO4•– are determined to be 0.43 ± 0.1 and 0.45 ± 0.1 per mol of O3 consumption, respectively. An adduct, –O3SOO– + O3 → –O3SO5–, is assumed as the first step, which further decomposes into SO5•– and O3•–. The subsequent reaction of SO5•– with O3 is proposed to generate SO4•–, while O3•– converts to •OH. A definition of Rct,•OH and Rct,SO4•– (i.e., respective ratios of •OH and SO4•– exposures to O3 exposure) is ...

Bit-Scalable Deep Hashing With Regularized Similarity Learning for Image Retrieval and Person Re-Identification

Abstract: Extracting informative image features and learning effective approximate hashing functions are two crucial steps in image retrieval. Conventional methods often study these two steps separately, e.g., learning hash functions from a predefined hand-crafted feature space. Meanwhile, the bit lengths of output hashing codes are preset in the most previous methods, neglecting the significance level of different bits and restricting their practical flexibility. To address these issues, we propose a supervised learning framework to generate compact and bit-scalable hashing codes directly from raw images. We pose hashing learning as a problem of regularized similarity learning. In particular, we organize the training images into a batch of triplet samples, each sample containing two images with the same label and one with a different label. With these triplet samples, we maximize the margin between the matched pairs and the mismatched pairs in the Hamming space. In addition, a regularization term is introduced to enforce the adjacency consistency, i.e., images of similar appearances should have similar codes. The deep convolutional neural network is utilized to train the model in an end-to-end fashion, where discriminative image features and hash functions are simultaneously optimized. Furthermore, each bit of our hashing codes is unequally weighted, so that we can manipulate the code lengths by truncating the insignificant bits. Our framework outperforms state-of-the-arts on public benchmarks of similar image search and also achieves promising results in the application of person re-identification in surveillance. It is also shown that the generated bit-scalable hashing codes well preserve the discriminative powers with shorter code lengths.

Constructing Uniform Core-Shell PPy@PANI Composites with Tunable Shell Thickness toward Enhancement in Microwave Absorption.

Abstract: Highly uniform core–shell composites, polypyrrole@polyaniline (PPy@PANI), have been successfully constructed by directing the polymerization of aniline on the surface of PPy microspheres. The thickness of PANI shells, from 30 to 120 nm, can be well controlled by modulating the weight ratio of aniline and PPy microspheres. PPy microspheres with abundant carbonyl groups have very strong affinity to the conjugated chains of PANI, which is responsible for the spontaneous formation of uniform core–shell microstructures. However, the strong affinity between PPy microspheres and PANI shells does not promote the diffusion or reassembly of two kinds of conjugated chains. Coating PPy microspheres with PANI shells increases the complex permittivity and creates the mechanism of interfacial polarization, where the latter plays an important role in increasing the dielectric loss of PPy@PANI composites. With a proper thickness of PANI shells, the moderate dielectric loss will produce well matched characteristic impedanc...

Ultrathin pancharatnam-berry metasurface with maximal cross-polarization efficiency.

Abstract: Novel ultrathin dual-functional metalenses are proposed, fabricated, tested, and verified in the microwave regime for the first time. The significance is that their anomalous transmission efficiency almost reaches the theoretical limit of 25%, showing a remarkable improvement compared with earlier ultrathin metasurface designs with less than 5% coupling efficiency. The planar metalens proposed empowers significant reduction in thickness, versatile focusing behavior, and high transmission efficiency simultaneously.

Assessment of Microbial Fuel Cell Configurations and Power Densities

Abstract: Different microbial electrochemical technologies are being developed for many diverse applications, including wastewater treatment, biofuel production, water desalination, remote power sources, and biosensors. Current and energy densities will always be limited relative to batteries and chemical fuel cells, but these technologies have other advantages based on the self-sustaining nature of the microorganisms that can donate or accept electrons from an electrode, the range of fuels that can be used, and versatility in the chemicals that can be produced. The high cost of membranes will likely limit applications of microbial electrochemical technologies that might require a membrane. For microbial fuel cells, which do not need a membrane, questions about whether larger-scale systems can produce power densities similar to those obtained in laboratory-scale systems remain. It is shown here that configuration and fuel (pure chemicals in laboratory media vs actual wastewaters) remain the key factors in power pro...

Facile synthesis of electrospun MFe2O4 (M = Co, Ni, Cu, Mn) spinel nanofibers with excellent electrocatalytic properties for oxygen evolution and hydrogen peroxide reduction

Abstract: Designing and preparing porous transition metal ferrites without using any template, shape-directing agent, and surfactant is a challenge. Herein, heterojunction MFe2O4 (M = Co, Ni, Cu, Mn) nanofiber (NF) based films with three-dimensional configurations were synthesized by electrospinning and the subsequent thermal treatment processes. Characterization results indeed show the 3D net-like textural structures of the electrospun spinel-type MFe2O4 NFs. In particular, the resulting MFe2O4 NFs have lengths up to several dozens of micrometers with an average diameter size of about 150 nm and possess abundant micro/meso/macropores on both the surface and within the films. The hierarchically porous structures and high surface areas of these MFe2O4 NFs (for example, the CoFe2O4 NFs possess a larger BET specific surface area (61.48 m2 g−1) than those of the CoFe2O4 NPs (5.93 m2 g−1)) can afford accessible transport channels for effectively decreasing the mass transport resistances, enhancing the electrical conductivity, and increasing the density and reactivity of the exposed catalytic active sites. All these advantages will be responsible for the better electrocatalytic performances of these MFe2O4 NFs compared with their structural isomers (i.e. the MFe2O4 NPs) for the oxygen evolution reaction (OER) and H2O2 reduction in alkaline solution. Meanwhile, both the OER and H2O2 reduction catalytic activities for these MFe2O4 NFs obey the order of CoFe2O4 NFs > CuFe2O4 NFs > NiFe2O4 NFs > MnFe2O4 NFs > Fe2O3 NFs. The CoFe2O4 NFs represent a new class of highly efficient non-noble-metal catalysts for both OER and H2O2 reduction/detection in alkaline media.

Formation of Uniform Fe3O4 Hollow Spheres Organized by Ultrathin Nanosheets and Their Excellent Lithium Storage Properties

Abstract: Hierarchical Fe3 O4 hollow spheres constructed by nanosheets are obtained from solvothermally synthesized Fe-glycerate hollow spheres. With the unique structural features, these hierarchical Fe3 O4 hollow spheres exhibit excellent electrochemical lithium-storage performance.

Patch Group Based Nonlocal Self-Similarity Prior Learning for Image Denoising

Abstract: Patch based image modeling has achieved a great success in low level vision such as image denoising. In particular, the use of image nonlocal self-similarity (NSS) prior, which refers to the fact that a local patch often has many nonlocal similar patches to it across the image, has significantly enhanced the denoising performance. However, in most existing methods only the NSS of input degraded image is exploited, while how to utilize the NSS of clean natural images is still an open problem. In this paper, we propose a patch group (PG) based NSS prior learning scheme to learn explicit NSS models from natural images for high performance denoising. PGs are extracted from training images by putting nonlocal similar patches into groups, and a PG based Gaussian Mixture Model (PG-GMM) learning algorithm is developed to learn the NSS prior. We demonstrate that, owe to the learned PG-GMM, a simple weighted sparse coding model, which has a closed-form solution, can be used to perform image denoising effectively, resulting in high PSNR measure, fast speed, and particularly the best visual quality among all competing methods.

A three-dimensional porous LiFePO4 cathode material modified with a nitrogen-doped graphene aerogel for high-power lithium ion batteries

Abstract: A composite cathode material consisting of (010) facet-oriented LiFePO4 nanoplatelets wrapped in a nitrogen-doped graphene aerogel is reported. Such a composite possesses a 3D porous structure with a BET surface area as high as 199.3 m2 g−1. In this composite, the nitrogen-doped graphene aerogel combined with its interconnected porous networks provides pathways for rapid electron transfer and ion transport, while the thin LFP nanoplatelets with large (010) surface area enhance the active sites and shorten the Li+ diffusion distances. As a result, a high rate capability (78 mA h g−1 at 100 C) as well as a long life cycling stability (89% capacity retention over 1000 cycles at 10 C) are achieved.

A robust smart window: reversibly switching from high transparency to angle-independent structural color display.

Abstract: A smart window is fabricated from a composite consisting of elastomeric poly(dimethylsiloxane) embedded with a thin layer of quasi-amorphous silica nanoparticles. The smart window can be switched from the initial highly transparent state to opaqueness and displays angle-independent structural color via mechanical stretching. The switchable optical property can be fully recovered after 1000 stretching/releasing cycles.

Convolutional Sparse Coding for Image Super-Resolution

Abstract: Most of the previous sparse coding (SC) based super resolution (SR) methods partition the image into overlapped patches, and process each patch separately. These methods, however, ignore the consistency of pixels in overlapped patches, which is a strong constraint for image reconstruction. In this paper, we propose a convolutional sparse coding (CSC) based SR (CSC-SR) method to address the consistency issue. Our CSC-SR involves three groups of parameters to be learned: (i) a set of filters to decompose the low resolution (LR) image into LR sparse feature maps, (ii) a mapping function to predict the high resolution (HR) feature maps from the LR ones, and (iii) a set of filters to reconstruct the HR images from the predicted HR feature maps via simple convolution operations. By working directly on the whole image, the proposed CSC-SR algorithm does not need to divide the image into overlapped patches, and can exploit the image global correlation to produce more robust reconstruction of image local structures. Experimental results clearly validate the advantages of CSC over patch based SC in SR application. Compared with state-of-the-art SR methods, the proposed CSC-SR method achieves highly competitive PSNR results, while demonstrating better edge and texture preservation performance.