Optimal Representation Research Articles

Human Face Identification based on Optimal Sparse Features

Security of human being is an important aspect in the context of data communication. To maintain security, technology is being developed from alpha-numeric passwords to biometric scanners. Recent advancement in security is the user authentication using face recognition. But the flaws in existing face recognition systems are yet to be addressed. This paper discusses solutions to the issues encountered by face recognition systems. Sparsity based classification is performed in this work. This method can handle errors occurs due to compress in and occlusion in a robust manner. We suggest a comprehensive classification algorithm characterized by sparse representation and 1l -minimization. In this method, the feature points and selection of features are not critical. The effect of change in occlusion can be easily addressed by using this optimal sparse representation based classification (OSRC) algorithm.

One View Per City for Buildings Segmentation in Remote-Sensing Images via Fully Convolutional Networks: A Proof-of-Concept Study.

The segmentation of buildings in remote-sensing (RS) images plays an important role in monitoring landscape changes. Quantification of these changes can be used to balance economic and environmental benefits and most importantly, to support the sustainable urban development. Deep learning has been upgrading the techniques for RS image analysis. However, it requires a large-scale data set for hyper-parameter optimization. To address this issue, the concept of “one view per city” is proposed and it explores the use of one RS image for parameter settings with the purpose of handling the rest images of the same city by the trained model. The proposal of this concept comes from the observation that buildings of a same city in single-source RS images demonstrate similar intensity distributions. To verify the feasibility, a proof-of-concept study is conducted and five fully convolutional networks are evaluated on five cities in the Inria Aerial Image Labeling database. Experimental results suggest that the concept can be explored to decrease the number of images for model training and it enables us to achieve competitive performance in buildings segmentation with decreased time consumption. Based on model optimization and universal image representation, it is full of potential to improve the segmentation performance, to enhance the generalization capacity, and to extend the application of the concept in RS image analysis.

Models and methods for workforce planning under uncertainty: Optimizing U.S. Army cyber branch readiness and manning

This work examines the problem of optimally resourcing personnel for a new set of U.S. Army cyber-specialty career fields using a combination of personnel accessions and inter-career-field transfers that are limited to occur over a subset of periods early within a 30-year career cycle. This workforce planning problem is bounded by constraints respectively pertaining to organizational needs for personnel, as well as personnel promotion policies. Complicating the problem are stochastic retention rates for every combination of year and promotion level within each career field over the 30-year period. Upon formalizing this Cyber Workforce Planning Problem (CWPP) within the framework of multiple performance goals, this study formulates, tests, and compares three frameworks to seek optimal workforce planning decisions under uncertainty: two stochastic programming (SP) variants and a robust optimization (RO) representation. Upon sampling the stochastic parametric distributions to generate a set of collectively representative, deterministic scenarios, the SP variants respectively leverage a scenario-based Monte Carlo approach and sample average approximation, whereas the RO model examines robust solutions over a range of decision-maker risk attitudes. A sensitivity analysis applied to the first of the SP variants indicates the solutions are relatively insensitive to different prioritizations of goals, and identifies policy insights to help recruit and retain personnel. Comparative testing of the three methodologies yields workforce planning recommendations that are relatively consistent across solution methodologies and identify concerns to inform changes to personnel policies.

Deep Latent Spectral Representation Learning-Based Hyperspectral Band Selection for Target Detection

Hyperspectral images (HSIs) can provide discriminative spectral signatures regarding the physical nature of different materials. It is this unique nature that makes HSIs to be of great interest in many fields. However, HSI application faces various challenges due to high dimensionality, redundant information, noisy bands, and insufficient samples. To address these problems, we propose an unsupervised band selection method based on deep latent spectral representation learning, called DLSRL, in this article. It imposes spectral consistency on deep latent space that resolves the issue of insufficient samples and spectral information lost in HSI interpretation. It pursues the low-dimensional optimal representation of the high-dimensional HSIs. In particular, an adaptive mapping relationship is constructed between the deep latent representation and the optimal subset to preserve physical significance optimally. Furthermore, a hierarchical optimization approach is introduced to achieve target detection with the selected subset. To verify the superiority of the proposed method, experiments have been conducted on four data sets captured by different sensors over different scenes. Comparative analyses validate that the proposed method presents superior performance in terms of high detection accuracy and low false alarm rate.

Exposure Measurement and Fusion via Adaptive Multiscale Edge-Preserving Smoothing

Natural environments usually have a larger dynamic range than the dynamic range that can be acquired by an optical camera with a single shot. In this paper, we propose a multiexposure fusion method that effectively fuses in a direct manner differently exposed images of a high dynamic range scene into a high-quality image. First, we present a developed joint weight by considering the exposure level measurement of the local and global luminance components of the input images. Second, we introduce a designed multiscale edge-preserving smoothing (MEPS) model for direct representing the weight maps. Third, two scale-aware factors for the MEPS model are adaptively determined without manual interference to obtain an optimal representation effect for each scale of the weight maps. The proposed adaptive MEPS model does not require Gaussian filtering steps to first smoothen the weight maps. It significantly reduces spatial artifacts in the fused image. We compare the proposed method with eight existing methods on 30 sequences from two databases with different characteristics. The experimental results indicate that the proposed approach achieves better imaging performance than the existing state-of-the-art methods, on both quantitative and qualitative evaluation. Moreover, it maintains a high computational efficiency.

Multi-Objective Sparse Subspace Clustering for Hyperspectral Imagery

Hyperspectral images (HSIs) are typical high-dimensional and complex data. As such, the clustering of HSIs is a challenging task. Out of the motivation to find the low-dimensional structure representation of the high-dimensional data, sparse subspace clustering (SSC) methods have been proposed in recent studies. Sparse representation is an important technique in SSC, which is aimed at obtaining the sparse coefficient matrix of the HSI data. Generally speaking, the acquisition of the sparse coefficient matrix is an ill-posed problem, and the existing methods introduce an extra condition as a regularization term to resolve it. However, the regularization parameter is determined manually, which is difficult and lacks self-adaptability. Hence, in this article, a multi-objective SSC method for hyperspectral imagery is proposed, which simultaneously optimizes the sparse term and the data fidelity term. In addition, the spatial structure information of the HSIs is often neglected in the processing model, and thus, a spatial prior term, as the third optimization objective function, is also tested in this article. As a result, there is no need to manually set a regularization parameter. Furthermore, by using the l <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sub> norm as the sparse term, this reduces the error caused by the convex relaxation of the other norms. In the proposed method, a multi-objective optimization model is first used to acquire the sparse coefficient matrix, in which a strategy for constructing the dictionary is proposed for more precise and efficient multi-objective optimization. In addition, a knee point-based selection method is utilized to automatically select the optimal sparse representation solution from the Pareto front. The adjacency matrix is then constructed according to the sparse coefficient matrix. Finally, a spectral clustering method is used to obtain clustering results. Experiments undertaken with four HSI data sets confirm the effectiveness of the proposed method.

Multiexposure Estimation and Fusion Based on a Sparsity Exposure Dictionary

Multiexposure image fusion (MEF), which is an emerging hot research topic, aims to fuse multiexposure images into a high-quality low dynamic range (LDR) image. In this article, we propose a novel exposure estimation-based MEF method with sparse decomposition and a designed sparsity exposure dictionary (SED). First, in order to represent more elementary information of multiexposure images, SED is trained by underexposed and overexposed images and constructed by the K-means-based algorithm (K-SVD) algorithm. Next, the SED is applied to construct exposure estimation maps by the number of atoms of the image patch, which is obtained by sparse decomposition. When the exposure estimation maps are obtained, a novel fusion framework that uses these maps to fuse multiexposure images is presented. In this framework, we employ the exposure estimation maps and an adaptive guided filter to construct the final fusion decision maps. The two factors of guided filter are adaptively determined without manual interference to obtain an optimal representation effect for multiexposure image sequences with different sizes and number. Finally, a fused image is generated via the pyramid merging method. The proposed fusion method is conducted on a series of multiexposure image sequences and compared to seven popular MEF methods. The experimental results clearly demonstrate that the proposed method outperforms the state-of-the-art methods in terms of subjective visual and quantitative evaluations.

Spectrally resolved cosmic ray hydrodynamics – I. Spectral scheme

Abstract Cosmic ray (CR) protons are an important component in many astrophysical systems. Processes like CR injection, cooling, adiabatic changes as well as active CR transport through the medium strongly modify the CR momentum distribution and have to be taken into account in hydrodynamical simulations. We present an efficient novel numerical scheme to accurately compute the evolution of the particle distribution function by solving the Fokker-Planck equation with a low number of spectral bins (10 − 20), which is required to include a full spectrum for every computational fluid element. The distribution function is represented by piecewise power laws and is not forced to be continuous, which enables an optimal representation of the spectrum. The Fokker-Planck equation is solved with a two-moment approach evolving the CR number and energy density. The low numerical diffusion of the scheme reduces the numerical errors by orders of magnitude in comparison to classical schemes with piecewise constant spectral representations. With this method not only the spectral evolution of CRs can be computed accurately in magnetohydrodynamic simulations but also their dynamical impact as well as CR ionisation. This allows for more accurate models for astrophysical plasmas, like the interstellar medium, and direct comparisons with observations.

Modeling Brain Diverse and Complex Hemodynamic Response Patterns via Deep Recurrent Autoencoder

For decades, a variety of task-based functional MRI (tfMRI) data analysis approaches have been developed, including the general linear model (GLM), sparse representations, and independent component analysis (ICA). However, these methods are mainly shallow models and are limited in faithfully modeling the complex, diverse, and concurrent spatial–temporal functional brain activities. Recently, recurrent neural networks (RNNs) have demonstrated great superiority in modeling temporal dependency of signals, while autoencoder models have been proven to be effective in automatically estimating the optimal representations of the original data. These characteristics of RNNs and autoencoders naturally meet the requirement of modeling hemodynamic response patterns in tfMRI data. Thus, we propose a novel unsupervised framework of deep recurrent autoencoder (DRAE) for modeling hemodynamic response patterns in this article. The basic idea of the DRAE model is to combine the deep RNN and the autoencoder to automatically characterize the meaningful functional brain networks and corresponding diverse and complex hemodynamic response patterns simultaneously. The experimental results demonstrate the superiority of the proposed DRAE model in automatically estimating the diverse and complex hemodynamic response patterns.

An Efficient and Accurate GB-SAR Imaging Algorithm Based on the Fractional Fourier Transform

In this paper, an efficient and accurate imaging algorithm is presented for Ground-Based Synthetic Aperture Radar (GB-SAR) or other radar systems that could be formed by a physical or synthetic linear aperture. The imaging algorithm is based on the fractional Fourier transform (FrFT) for the azimuth compression. A mathematical framework is derived according to the projection of a sample reflectivity image onto the pseudopolar coordinate, and its implementation was presented. With the data acquisition geometry and the pseudopolar imaging coordinate, the phase of a point target can be expressed as a quadratic phase exponential. It makes that only 1-D FrFT is needed for the azimuth compression of the time-domain backscatter data for the GB-SAR imaging problem. By further research, the optimal transformation order that represents the spatial frequency changes by the FrFT was given subsequently. Taking advantage of this optimal representation, the proposed approach avoids the large calculation that occurs in the time-domain back projection (TDBP). Comparing to the far-field pseudopolar format algorithm (FPFA), the accuracy of the proposed algorithm is much improved. Meanwhile, the proposed approach holds almost the same computational cost and complexity as the FPFA. The proposed approach keeps the advantages of the imaging quality of the TDBP and the computational cost of the FPFA that are two important aspects of the GB-SAR applications. Both the numerical simulation and the field GB-SAR experiment show that the algorithm is more suitable for the high-precision GB-SAR imaging, especially for the near field.

Linear Discriminative Learning for Image Classification

In this paper, we propose a linear discriminative learning model called adaptive locality-based weighted collaborative representation (ALWCR) that formulates the image classification task as an optimization problem to reduce the reconstruction error between the query sample and its computed linear representation. The optimal linear representation for a query image is obtained by using the weighted regularized linear regression approach which incorporates intrinsic locality structure and feature variance between data into representation. The resultant representation increases the discrimination ability for correct classification. The proposed ALWCR method can be considered an extension of the collaborative representation- (CR-) based classification approach which is an alternative to the sparse representation- (SR-) based classification method. ALWCR improved the discriminant ability for classification as compared with CR original formulation and overcomes the limitations that arose due to a small training sample size and low feature dimension. Experimental results obtained using various feature dimensions on well-known publicly available face and digit datasets have verified the competitiveness of the proposed method against competing image classification methods.

Unsupervised Multi-Label Image and Texture Segmentation based on Optimal Feature Representation

Unsupervised Multi-Label Image and Texture Segmentation based on Optimal Feature Representation

Determining Optimal Coarse-Grained Representation for Biomolecules Using Internal Cluster Validation Indexes.

The development of ultracoarse-grained models for large biomolecules needs to derive the optimal number of coarse-grained (CG) sites to represent the targets. In this work, we propose to use the statistical internal cluster validation indexes to determine the optimal number of CG sites that are optimized based on the essential dynamics coarse-graining method. The calculated curves of Calinski-Harabasz and Silhouette Coefficient indexes exhibit the extrema corresponding to the similar CG numbers. The calculated ratios of the optimal CG numbers to the residue numbers of fine-grained models are in the range from 4 to 2. The comparison of the stability of index results indicates that Calinski-Harabasz index is the better choice to determine the optimal CG representation in coarse-graining. © 2019 Wiley Periodicals, Inc.

Simulation Tests of the Accuracy of Fitting Two Freeform Surfaces

The presented method is introduced to simulate and predict the accuracy of fitting two freeform surfaces. For this purpose, the CAD models of both actual surfaces should be determined on the basis of coordinate measurement data obtained in measurements along regular grids of points in the UV space. The NURBS regression surfaces are modeled on the measurement data. Adequate regression models are sought with the iterative procedure. In the following steps of the procedure, the number of control points and/or the degree of the surface is/are changed, and the autocorrelation of residuals from the models is tested using the spatial statistics methods. The designated models are optimal CAD representations of the actual surfaces. Tests of the accuracy of fitting the surfaces are carried out virtually by fitting together both models in the CAD software. The outcome of the study is a spatial model of the gap between the studied surfaces. The obtained model was verified experimentally by measuring the dimensions of the actual gap between the surfaces, applying a measuring microscope. The proposed method is a useful tool in analyzing and improving the accuracy of injection molds machining.

Image Fusion Technique Based on Hybrid Whale Optimization Algorithm Simulated Annealing (hWOA-SA)

Curvelet transform is a multiscale directional transformer, which allows optimal non-adaptive sparse representation of object with edge. In this paper, a new image fusion technique has been developed by combination of whale optimization algorithm (WOA) and simulated annealing (SA) along with curvelet transform. The resulting combined algorithm is abbreviated as hybrid whale optimization algorithm with simulated annealing. Initially, hWOA-SA has been applied to enhancing the quality of image using de-noising scheme. Afterwards, the curvelet transform has been employed to carry out the fusion of images. In terms of PSNR, the curvelet transform exhibits the better performance. The effectiveness and validation of the proposed scheme has been carried-out using quality matrices. The performance analysis is carried out after checking the effectiveness of proposed approach by evaluating the various parameters such as: RSME, PFE, MAE, CORR, SNR, PSNR, MI, UQI and SSIM and compared with numerous techniques. Simulation results obtained from proposed hWOA-SA based image fusion are very competitive and better than other image fusion technique available in the literature.

A structural property of the wavelet packet transform method to localise incoherency of a signal

This paper proves a general structural property of the wavelet tree for a given seminorm in the context of the wavelet packet transform method. This structural property can be used in denoising algorithms of different applications to guarantee the optimality of novel search strategies. The property holds for any input signals using any orthogonal analysing wavelet families. The property holds for any norms, which results to be a convex function through the wavelet tree. Using a defined norms, seminorms or pseudonorms, this property can be used to detect incoherent parts of an input signal by using the minimal depth of the tree. In this sense the proposed denoising procedure works without thresholds for the localisation of different kinds of noise, as well as for a stop criterium for an optimal representation of the incoherency. The proof of this property is performed by mathematical induction and the demonstration is based on orthonormality with the help of the multiresolution framework aspects of the wavelets packet tree. The Theorem is independent of the definition of the adopted norm and of the incoherent part of an input signal. In this sense, the discovered property is a general one, which is related to any norm and any nature of the signal incoherence. It can be used in different applications, in which a minimum of a norm is required to be calculated through the wavelet tree.

Optimizing sampling strategy for Chinese National Sewage Sludge Survey (CNSSS) based on urban agglomeration, wastewater treatment process, and treatment capacity

As a sink and source of contaminants, sewage sludge is a good matrix to capture the spatial-temporal trend of chemicals and assess the potential risks these chemicals pose to human health and the environment. In order to understand these chemical risks, a robust statistical sewage sludge sampling strategy for Chinese wastewater treatment plants (WWTPs) must be designed. The purpose of this paper is to develop such a sampling strategy for Chinese WWTPs which may be used optimally. Before creating the sampling design, the distribution of WWTPs was categorically analyzed. These categories include urban agglomeration, wastewater treatment process, and wastewater treatment capacity. Particular attention was given to the studying of population distribution, gross domestic product, WWTP number, wastewater treatment flow, and dry sludge production in each urban agglomeration. In addition, correlation analysis was conducted among these five indexes. Due to the heterogeneity of WWTPs, stratified sampling had to be used to homogenize the sampling units. The eight strategies proposed herein were based on simple random sampling and stratified random sampling methods. Moreover, the aforementioned three categories (urban agglomeration, treatment process, and treatment capacity) were intended to be stratification indicators. Furthermore, Monte Carlo simulations revealed that the treatment capacity based stratified random sampling strategy (Strategy 4) results in the optimal sample representation, with the smallest root mean square error compared to seven other sampling strategies with different strata. This optimal stratified sampling strategy, if employed during the Chinese national sewage sludge survey, has the potential to greatly contribute to data quality and assurance.

Statistical analysis for satellite-index-based insurance to define damaged pasture thresholds

Abstract. Vegetation indices based on satellite images, such as the normalized difference vegetation index (NDVI), have been used in countries like the USA, Canada and Spain for damaged pasture and forage insurance over the last few years. This type of agricultural insurance is called satellite-index-based insurance (SIBI). In SIBI, the occurrence of damage is defined as normal distributions. In this work a pasture area at the north of the Community of Madrid (Spain) has been delimited by means of Moderate Resolution Imaging Spectroradiometer (MODIS) images. A statistical analysis of NDVI histograms was applied to seek for alternative distributions using the maximum likelihood method and χ2 test. The results show that the normal distribution is not the optimal representation and the generalized extreme value (GEV) distribution presents a better fit through the year based on a quality estimator. A comparison between normal and GEV is shown with respect to the probability under a NDVI threshold value throughout the year. This suggests that an a priori distribution should not be selected and a percentile methodology should be used to define a NDVI damage threshold rather than the average and standard deviation, typically of normal distributions. Highlights. The GEV distribution provides better fit to the NDVI historical observations than the normal one. Differences between normal and GEV distributions are higher during spring and autumn, which are transition periods in the precipitation regimen. NDVI damage threshold shows evident differences using normal and GEV distributions both covering the same probability (24.20 %). NDVI damage threshold values based on percentile calculation are proposed as an improvement in the index-based insurance in damaged pasture.

Segmentation of bias field induced brain MR images using rough sets and stomped-t distribution

Over the past few decades, automatic segmentation of brain magnetic resonance (MR) images into different tissue classes has remained an important research area, particularly due to the presence of bias field artifact in MR images. In this regard, the stomped normal (StN) distribution is proved to generate an optimal representation of the intensity distribution in brain MR images, by incorporating the properties of rough sets in the probabilistic framework. The StN distribution is capable of successfully modelling the central tendency, dispersion, and width of the intensity distribution. However, it does not take into consideration the kurtosis of the distribution, which controls the concentration of values around the mean and shape of the tail of intensity distribution. In this regard, the paper presents a novel method for simultaneous segmentation and bias field correction in brain MR images. It integrates the concept of rough sets and the merit of a recently introduced probability distribution, called stomped-t (St-t) distribution. The St-t distribution incorporates the property of kurtosis in rough-probabilistic framework, where each tissue class is modelled using a crisp lower approximation and a probabilistic boundary region. The brain MR image is modelled using a mixture of finite number of St-t distributions and one uniform distribution. The uniform distribution takes into account cerebro-spinal fluid, pathologies, and other non-brain tissues. The proposed method employs both expectation-maximization algorithm and hidden Markov random field model for accurate and robust image segmentation. The performance of the proposed approach, along with a comparison with related methods, is illustrated on a set of synthetic and real brain MR images for different bias fields and noise levels.

The Use of Deep Learning Distributed Representations in the Identification of Abusive Text

The selection of optimal feature representations is a critical step in the use of machine learning in text classification. Traditional features (e.g. bag of words and n-grams) have dominated for decades, but in the past five years, the use of learned distributed representations has become increasingly common. In this paper, we summarise and present a categorisation of the stateof-the-art distributed representation techniques, including word and sentence embedding models. We carry out an empirical analysis of the performance of the various feature representations using the scenario of detecting abusive comments. We compare classification accuracies across a range of off-the-shelf embedding models using 10 labelled datasets gathered from different social media platforms. Our results show that multi-task sentence embedding models perform best with consistently highest classification results in comparison to other embedding models. We hope our work can be a guideline for practitioners in selecting appropriate features in text classification task, particularly in the domain of abuse detection.