Search results

This paper aims to present a novel approach of image super-resolution based on deep–shallow cascaded convolutional neural networks for reconstructing a clear and high-resolution (HR) remote sensing image from a low-resolution (LR) input.

The proposed approach directly learns the residuals and mapping between simulated LR and their corresponding HR remote sensing images based on deep and shallow end-to-end convolutional networks instead of assuming any specific restored models. Extra max-pooling and up-sampling are used to achieve a multiscale space by concatenating low- and high-level feature maps, and an HR image is generated by combining LR input and the residual image. This model ensures a strong response to spatially local input patterns by using a large filter and cascaded small filters. The authors adopt a strategy based on epochs to update the learning rate for boosting convergence speed.

The proposed deep network is trained to reconstruct high-quality images for low-quality inputs through a simulated dataset, which is generated with Set5, Set14, Berkeley Segmentation Data set and remote sensing images. Experimental results demonstrate that this model considerably enhances remote sensing images in terms of spatial detail and spectral fidelity and outperforms state-of-the-art SR methods in terms of peak signal-to-noise ratio, structural similarity and visual assessment.

The proposed method can reconstruct an HR remote sensing image from an LR input and significantly improve the quality of remote sensing images in terms of spatial detail and fidelity.

The purpose of this paper is to build a compact and accurate classifier for high-dimensional classification.

A classification approach based on class-dependent feature subspace (CFS) is proposed. CFS is a class-dependent integration of a support vector machine (SVM) classifier and associated discriminative features. For each class, our genetic algorithm (GA)-based approach evolves the best subset of discriminative features and SVM classifier simultaneously. To guarantee convergence and efficiency, the authors customize the GA in terms of encoding strategy, fitness evaluation, and genetic operators.

Experimental studies demonstrated that the proposed CFS-based approach is superior to other state-of-the-art classification algorithms on UCI data sets in terms of both concise interpretation and predictive power for high-dimensional data.

UCI data sets rather than real industrial data are used to evaluate the proposed approach. In addition, only single-label classification is addressed in the study.

The proposed method not only constructs an accurate classification model but also obtains a compact combination of discriminative features. It is helpful for business makers to get a concise understanding of the high-dimensional data.

The authors propose a compact and effective classification approach for high-dimensional data. Instead of the same feature subset for all the classes, the proposed CFS-based approach obtains the optimal subset of discriminative feature and SVM classifier for each class. The proposed approach enhances both interpretability and predictive power for high-dimensional data.

For an entrant platform in the on-demand service market, choosing an appropriate employment model is critical. This study explores how the entrant optimally chooses the employment model to achieve better performance and investigates the optimal pricing strategies and wage schemes for both incumbent and entrant platforms.

Based on the Hotelling model, the authors develop a game-theoretic framework to study the incumbent's and entrant's optimal service prices and wage schemes. Moreover, the authors determine the entrant's optimal employment model by comparing the entrant's optimal profits under different market configurations and analytically analyze the impacts of some critical factors on the platforms' decision-making.

This study reveals that the impacts of the unit misfit cost of suppliers or consumers on the pricing strategies and wage schemes vary with different operational efficiencies of platforms. Only when both the service efficiency of contractors and the basic employee benefits are low, entrants should adopt the employee model. Moreover, a lower unit misfit cost of suppliers or consumers makes entrants more likely to choose the contractor model. However, the service efficiency of contractors has nonmonotonic effects on the entrant's decision.

This study focuses on an entrant's decision on the optimal employment model in an on-demand service market, considering the competition between entrants and incumbents on both the supplier and consumer sides, which has not been investigated in the prior literature.

With the increasing heat dissipation of electronic devices, the cooling demand of electronic products is increasing gradually. A water-cooled microchannel heat exchanger is an effective cooling technology for electronic equipment. The structure of a microchannel has great impact on the heat transfer performance of a microchannel heat exchanger. The purpose of this paper is to analyze and compare the fluid flow and heat transfer characteristic of a microchannel heat exchanger with different reentrant cavities.

The three-dimensional steady, laminar developing flow and conjugate heat transfer governing equations of a plate microchannel heat exchanger are solved using the finite volume method.

At the flow rate range studied in this paper, the microchannel heat exchangers with reentrant cavities present better heat transfer performance and smaller pressure drop. A microchannel heat exchanger with trapezoidal-shaped cavities has best heat transfer performance, and a microchannel heat exchanger with fan-shaped cavities has the smallest pressure drop.

The fluid is incompressible and the inlet temperature is constant.

It is an effective way to enhance heat transfer and reduce pressure drop by adding cavities in microchannels and the data will be helpful as guidelines in the selection of reentrant cavities.

This paper provides the pressure drop and heat transfer performance analysis of microchannel heat exchangers with various reentrant cavities, which can provide reference for heat transfer augmentation of an existing microchannel heat exchanger in a thermal design.

The purpose of the article is to investigate the influence of deburring by wire brushing upon states of magnesium alloy edges.

AZ91HP and AZ31 magnesium alloy samples were machined with the use of recommended, catalog milling parameters. Burrs formed at the edges after milling were removed by brushing. Three different kinds of brushes were tested. Edge states (values of edge radius) after wire brushing were specified. Surface roughness was measured near the brushed edges.

Experimental results show that wire brushing is an efficient deburring method, which can be fully automated on machining centers. Depending on the requirements, specific values of edge radius as well as surface roughness may be obtained.

The article will help technological process designers select tools for deburring after milling of magnesium alloys.

The paper presents the automated deburring method which can provide the required edge radius of aerospace components.

In this study, we propose a tetrahedral mesh generation and adaptive refinement method for multi-chamber, multi-facet, multiscale and surface-solid mesh coupling with extremely thin layers, solving the two challenges of mesh generation and refinement in current electromagnetic simulation models.

Utilizing innovative topology transformation techniques, high-precision intersection judgment algorithms and highly reliable boundary recovery algorithms to reduce the number of Steiner locking points. The feasible space for the reposition of Steiner points is determined by using the linear programming. During mesh refinement, an edge-split method based on geometric center and boundary facets node size is devised. Solving the problem of difficult insertion of nodes in narrow geometric spaces, capable of filtering the longest and boundary edges of tetrahedrons, refining the mesh layer by layer through multiple iterations, and achieving collaborative optimization of surface and tetrahedral mesh. Simultaneously, utilizing a surface-facet preserving mesh topology optimization algorithm to improve the fit degree between the mesh and geometry.

Initial mesh generation for electromagnetic models, compared to commercial software, the method proposed in this paper has a higher pass rate and better mesh quality. For the adaptive refinement performance of high-frequency computing, this method can generate an average of 50% fewer meshes compared to commercial software while meeting simulation accuracy.

This paper proposes a complete set of mesh generation and adaptive refinement theories and methods designed for the structural characteristics of electromagnetic simulation models, which meet the needs of real-world industrial applications.