Sparse code shrinkage: denoising of nongaussian data by maximum likelihood estimation

An enhanced K-SVD denoising algorithm based on adaptive soft-threshold shrinkage for fault detection of wind turbine rolling bearing

Due to nonstationary operating conditions of wind turbines and surrounding harsh working environments, the impulse features induced by bearing faults are always overwhelmed by heavy noise, which brings challenges to accurately detect rolling bearing faults. Sparse representation exhibits excellent performance in nonstationary signal analysis, but it is closely bound up with the degree of similarity between the atoms in a dictionary and signals. Therefore, this paper investigates an enhanced K-SVD denoising method based on adaptive soft-threshold shrinkage to achieve high-precision extraction of impulse signals, and applies it to fault detection of generator bearing of wind turbines. An adaptive sparse coding shrinkage soft-threshold denoising is first proposed to remove noise and harmonic interference in the residual term of dictionary updating, so that the updated atoms show obvious impact characteristics. Furthermore, a soft-threshold shrinkage function with adaptive threshold is designed to further suppress clutter in atoms of the learned dictionary, so as to obtain an optimized dictionary for recovering impulse signals. Two actual engineering cases are selected for analysis, and the envelope spectrum correlation kurtosis corresponding to the results obtained by the proposed method is significantly higher than that of other comparison methods, thus verifying its superiority in detecting rolling bearing faults.

A Novel Joint Channel Estimation and Symbol Detection Receiver for Orthogonal Time Frequency Space in Vehicular Networks

A vehicular network embodies a specialized variant of wireless network systems, characterized by its capability to facilitate inter-vehicular communication and connectivity with the encompassing infrastructure. With the rapid development of wireless communication technology, high-speed and reliable communication has become increasingly important in vehicular networks. It has been demonstrated that orthogonal time frequency space (OTFS) modulation proves effective in addressing the challenges posed by high-mobility environments, as it transforms the time-varying channels into the delay-Doppler domain. Motivated by this, in this paper, we focus on the theme of integrated sensing and communication (ISAC)-assisted OTFS receiver design, which aims to perform sensing channel estimation and communication symbol detection. Specifically, the estimation of the sensing channel is accomplished through the utilization of a deep residual denoising network (DRDN), while the communication symbol detection is performed by orthogonal approximate message passing (OAMP) processing. The numerical results demonstrate that the proposed ISAC system exhibits superior performance and robustness compared to traditional methods, with a lower complexity as well. The proposed system has great potential for future applications in wireless communication systems, especially in challenging scenarios with high mobility and interference.

Rotating machine fault diagnosis by a novel fast sparsity-enabled feature-energy-ratio method

To well extract the early fault characteristics of rotating machines, a new fast sparsity-enabled feature-energy-ratio method is investigated in this paper. This method includes two stages. In the first stage, the spectrum is adaptively segmented through a coarse-to-fine strategy based on the ordered local maximums. Thus, the fault characteristic band can be divided automatically. A novel index based on sparsity, energy ratio, and kurtosis, is constructed to evaluate periodic impulses in each sub-signal, and it can evaluate the periodic impulses from the globality and locality. In the second stage, the Fourier spectrum from the first stage are refined by an improved sparse coding shrinkage denoising (SCSD) method whose parameters can be dynamically determined for each point. Within the improved SCSD approach, the differential result of the amplitude spectrum is used as input to improve the sparsity. Moreover, the ratios between the SCSD output and its input are applied to weigh the Fourier spectrum and maintain the phase information. Finally, the inverse fast Fourier transform and squared envelope spectra are applied to detect the fault characteristics. Bearing and gearbox vibration signals are used to validate the proposed methodology. The experimental results show that the proposed method is superior to some typical methods and the proposed index are robust to the interferences from aperiodic impulses. Therefore, the proposed method has great potential in the fault diagnosis of rotating machine.

Efficient coding theory of dynamic attentional modulation

Activity of sensory neurons is driven not only by external stimuli but also by feedback signals from higher brain areas. Attention is one particularly important internal signal whose presumed role is to modulate sensory representations such that they only encode information currently relevant to the organism at minimal cost. This hypothesis has, however, not yet been expressed in a normative computational framework. Here, by building on normative principles of probabilistic inference and efficient coding, we developed a model of dynamic population coding in the visual cortex. By continuously adapting the sensory code to changing demands of the perceptual observer, an attention-like modulation emerges. This modulation can dramatically reduce the amount of neural activity without deteriorating the accuracy of task-specific inferences. Our results suggest that a range of seemingly disparate cortical phenomena such as intrinsic gain modulation, attention-related tuning modulation, and response variability could be manifestations of the same underlying principles, which combine efficient sensory coding with optimal probabilistic inference in dynamic environments.

Dual Mixture Model Based CNN for Image Denoising

Non-Gaussian residual error and noise are common in the real applications, and they can be efficiently addressed by some non-quadratic fidelity terms in the classic variational method. However, they have not been well integrated into the architectures design in the convolutional neural networks (CNN) based image denoising method. In this paper, we propose a deep learning approach to handle non-Gaussian residual error. Our method is developed on an universal approximation property for the probability density functions of the non-Gaussian error/noise. By considering the duality of the maximum likelihood estimation for the non-Gaussian error, an adaptive weighting strategy can be derived for image fidelity. To get a good image prior, a learnable regularizer is adopted. Solving such a problem iteratively can be unrolled as a weighted residual CNN architecture. The main advantage of our method is that the weighted residual block can well handle the non-Gaussian residual, especially for the noise with non-uniformly spatial distribution. Numerical results show that it has better performance on non-Gaussian noise (e.g. Gaussian mixture, random-valued impulse noise) removal than the related existing methods.

Noise reduction by adaptive-SIN filtering for retinal OCT images

Optical coherence tomography (OCT) images is widely used in ophthalmic examination, but their qualities are often affected by noises. Shearlet transform has shown its effectiveness in removing image noises because of its edge-preserving property and directional sensitivity. In the paper, we propose an adaptive denoising algorithm for OCT images. The OCT noise is closer to the Poisson distribution than the Gaussian distribution, and shearlet transform assumes additive white Gaussian noise. We hence propose a square-root transform to redistribute the OCT noise. Different manufacturers and differences between imaging objects may influence the observed noise characteristics, which make predefined thresholding scheme ineffective. We propose an adaptive 3D shearlet image filter with noise-redistribution (adaptive-SIN) scheme for OCT images. The proposed adaptive-SIN is evaluated on three benchmark datasets using quantitative evaluation metrics and subjective visual inspection. Compared with other algorithms, the proposed algorithm better removes noise in OCT images and better preserves image details, significantly outperforming in terms of both quantitative evaluation and visual inspection. The proposed algorithm effectively transforms the Poisson noise to Gaussian noise so that the subsequent shearlet transform could optimally remove the noise. The proposed adaptive thresholding scheme optimally adapts to various noise conditions and hence better remove the noise. The comparison experimental results on three benchmark datasets against 8 compared algorithms demonstrate the effectiveness of the proposed approach in removing OCT noise.

An enhanced rolling bearing fault detection method combining sparse code shrinkage denoising with fast spectral correlation

Rolling bearings are important supporting components widely used in rotating machinery and are prone to failure, it is thus important to perform fault detection of rolling bearing quickly and accurately. Aiming at the problem that it is difficult to extract the weak impulses buried in strong background noise in rolling bearing fault diagnosis, this paper proposes an enhanced fault detection method combining sparse code shrinkage denoising with fast spectral correlation according to the cyclic statistical properties of defective bearing vibration signals. First, in view of the non-Gaussian statistical properties of the periodic impulses caused by the localized bearing defect in vibration signals, the sparse code shrinkage algorithm is employed to denoise the original noisy signal, thereby highlighting the periodic impulses. Then, the Fast Spectral Correlation (Fast-SC) algorithm is used to process the denoised signal to get the cyclic spectral correlation. Finally, the squared enhanced envelope spectrum (SEES) is presented to effectively detect and identify the rolling bearing faults. Experimental results demonstrate the validity and superiority of the proposed method in rolling bearing fault detection through the comparison with the Fast-SC, spectral kurtosis and Infogram.

Regularized Group Sparse Discriminant Analysis for P300-Based Brain–Computer Interface

Event-related potentials (ERPs) especially P300 are popular effective features for brain–computer interface (BCI) systems based on electroencephalography (EEG). Traditional ERP-based BCI systems may perform poorly for small training samples, i.e. the undersampling problem. In this study, the ERP classification problem was investigated, in particular, the ERP classification in the high-dimensional setting with the number of features larger than the number of samples was studied. A flexible group sparse discriminative analysis algorithm based on Moreau–Yosida regularization was proposed for alleviating the undersampling problem. An optimization problem with the group sparse criterion was presented, and the optimal solution was proposed by using the regularized optimal scoring method. During the alternating iteration procedure, the feature selection and classification were performed simultaneously. Two P300-based BCI datasets were used to evaluate our proposed new method and compare it with existing standard methods. The experimental results indicated that the features extracted via our proposed method are efficient and provide an overall better P300 classification accuracy compared with several state-of-the-art methods.

An Early Fault Diagnosis Method of Rolling Bearings on the Basis of Adaptive Frequency Window and Sparse Coding Shrinkage

Early fault information of rolling bearings is weak and often submerged by background noise, easily leading to misdiagnosis or missed diagnosis. In order to solve this issue, the present paper puts forward a fault diagnosis method on the basis of adaptive frequency window (AFW) and sparse coding shrinkage (SCS). The proposed method is based on the idea of determining the resonance frequency band, extracting the narrowband signal, and envelope demodulating the extracted signal. Firstly, the paper introduces frequency window, which can slip on the frequency axis and extract the frequency band. Secondly, the double time domain feature entropy is proposed to evaluate the strength of periodic components in signal. The location of the optimal frequency window covering the resonance band caused by bearing fault is determined adaptively by this entropy index and the shifting/expanding frequency window. Thirdly, the signal corresponding to the optimal frequency window is reconstructed, and it is further filtered by the sparse coding shrinkage algorithm to highlight the impact feature and reduce the residue noise. Fourthly, the de-noised signal is demodulated by envelope operation, and the corresponding envelope spectrum is calculated. Finally, the bearing failure type can be judged by comparing the frequency corresponding to the spectral lines with larger amplitude in the envelope spectrum and the fault characteristic frequency. Two bearing vibration signals are applied to validate the proposed method. The analysis results illustrate that this method can extract more failure information and highlight the early failure feature. The data files of Case Western Reserve University for different operation conditions are used, and the proposed approach achieves a diagnostic success rate of 83.3%, superior to that of the AFW method, SCS method, and Fast Kurtogram method. The method presented in this paper can be used as a supplement to the early fault diagnosis method of rolling bearings.

MetICA: independent component analysis for high-resolution mass-spectrometry based non-targeted metabolomics

BACKGROUND

Interpreting non-targeted metabolomics data remains a challenging task. Signals from non-targeted metabolomics studies stem from a combination of biological causes, complex interactions between them and experimental bias/noise. The resulting data matrix usually contain huge number of variables and only few samples, and classical techniques using nonlinear mapping could result in computational complexity and overfitting. Independent Component Analysis (ICA) as a linear method could potentially bring more meaningful results than Principal Component Analysis (PCA). However, a major problem with most ICA algorithms is the output variations between different runs and the result of a single ICA run should be interpreted with reserve.

RESULTS

ICA was applied to simulated and experimental mass spectrometry (MS)-based non-targeted metabolomics data, under the hypothesis that underlying sources are mutually independent. Inspired from the Icasso algorithm, a new ICA method, MetICA was developed to handle the instability of ICA on complex datasets. Like the original Icasso algorithm, MetICA evaluated the algorithmic and statistical reliability of ICA runs. In addition, MetICA suggests two ways to select the optimal number of model components and gives an order of interpretation for the components obtained.

CONCLUSIONS

Correlating the components obtained with prior biological knowledge allows understanding how non-targeted metabolomics data reflect biological nature and technical phenomena. We could also extract mass signals related to this information. This novel approach provides meaningful components due to their independent nature. Furthermore, it provides an innovative concept on which to base model selection: that of optimizing the number of reliable components instead of trying to fit the data. The current version of MetICA is available at https://github.com/daniellyz/MetICA.

Removal of Vesicle Structures From Transmission Electron Microscope Images

In this paper, we address the problem of imaging membrane proteins for single-particle cryo-electron microscopy reconstruction of the isolated protein structure. More precisely, we propose a method for learning and removing the interfering vesicle signals from the micrograph, prior to reconstruction. In our approach, we estimate the subspace of the vesicle structures and project the micrographs onto the orthogonal complement of this subspace. We construct a 2D statistical model of the vesicle structure, based on higher order singular value decomposition (HOSVD), by considering the structural symmetries of the vesicles in the polar coordinate plane. We then propose to lift the HOSVD model to a novel hierarchical model by summarizing the multidimensional HOSVD coefficients by their principal components. Along with the model, a solid vesicle normalization scheme and model selection criterion are proposed to make a compact and general model. The results show that the vesicle structures are accurately separated from the background by the HOSVD model that is also able to adapt to the asymmetries of the vesicles. This is a promising result and suggests even wider applicability of the proposed approach in learning and removal of statistical structures.

Blind source separation based x-ray image denoising from an image sequence

Blind source separation (BSS) based x-ray image denoising from an image sequence is proposed. Without priori knowledge, the useful image signal can be separated from an x-ray image sequence, for original images are supposed as different combinations of stable image signal and random image noise. The BSS algorithms such as fixed-point independent component analysis and second-order statistics singular value decomposition are used and compared with multi-frame averaging which is a common algorithm for improving image's signal-to-noise ratio (SNR). Denoising performance is evaluated in SNR, standard deviation, entropy, and runtime. Analysis indicates that BSS is applicable to image denoising; the denoised image's quality will get better when more frames are included in an x-ray image sequence, but it will cost more time; there should be trade-off between denoising performance and runtime, which means that the number of frames included in an image sequence is enough.

Generalized multiple kernel learning with data-dependent priors

Multiple kernel learning (MKL) and classifier ensemble are two mainstream methods for solving learning problems in which some sets of features/views are more informative than others, or the features/views within a given set are inconsistent. In this paper, we first present a novel probabilistic interpretation of MKL such that maximum entropy discrimination with a noninformative prior over multiple views is equivalent to the formulation of MKL. Instead of using the noninformative prior, we introduce a novel data-dependent prior based on an ensemble of kernel predictors, which enhances the prediction performance of MKL by leveraging the merits of the classifier ensemble. With the proposed probabilistic framework of MKL, we propose a hierarchical Bayesian model to learn the proposed data-dependent prior and classification model simultaneously. The resultant problem is convex and other information (e.g., instances with either missing views or missing labels) can be seamlessly incorporated into the data-dependent priors. Furthermore, a variety of existing MKL models can be recovered under the proposed MKL framework and can be readily extended to incorporate these priors. Extensive experiments demonstrate the benefits of our proposed framework in supervised and semisupervised settings, as well as in tasks with partial correspondence among multiple views.

Discrimination of tomatoes bred by spaceflight mutagenesis using visible/near infrared spectroscopy and chemometrics

Visible/near infrared spectroscopy (Vis/NIR) based on sensitive wavelengths (SWs) and chemometrics was proposed to discriminate different tomatoes bred by spaceflight mutagenesis from their leafs or fruits (green or mature). The tomato breeds were mutant M1, M2 and their parent. Partial least squares (PLS) analysis and least squares-support vector machine (LS-SVM) were implemented for calibration models. PLS analysis was implemented for calibration models with different wavebands including the visible region (400-700 nm) and the near infrared region (700-1000 nm). The best PLS models were achieved in the visible region for the leaf and green fruit samples and in the near infrared region for the mature fruit samples. Furthermore, different latent variables (4-8 LVs for leafs, 5-9 LVs for green fruits, and 4-9 LVs for mature fruits) were used as inputs of LS-SVM to develop the LV-LS-SVM models with the grid search technique and radial basis function (RBF) kernel. The optimal LV-LS-SVM models were achieved with six LVs for the leaf samples, seven LVs for green fruits, and six LVs for mature fruits, respectively, and they outperformed the PLS models. Moreover, independent component analysis (ICA) was executed to select several SWs based on loading weights. The optimal LS-SVM model was achieved with SWs of 550-560 nm, 562-574 nm, 670-680 nm and 705-71 5 nm for the leaf samples; 548-556 nm, 559-564 nm, 678-685 nm and 962-974 nm for the green fruit samples; and 712-718 nm, 720-729 nm, 968-978 nm and 820-830 nm for the mature fruit samples. All of them had better performance than PLS and LV-LS-SVM, with the parameters of correlation coefficient (rp), root mean square error of prediction (RMSEP) and bias of 0.9792, 0.2632 and 0.0901 based on leaf discrimination, 0.9837, 0.2783 and 0.1758 based on green fruit discrimination, 0.9804, 0.2215 and -0.0035 based on mature fruit discrimination, respectively. The overall results indicated that ICA was an effective way for the selection of SWs, and the Vis/NIR combined with LS-SVM models had the capability to predict the different breeds (mutant M1, mutant M2 and their parent) of tomatoes from leafs and fruits.

Evaluation of the sparse coding shrinkage noise reduction algorithm in normal hearing and hearing impaired listeners

Although there are numerous single-channel noise reduction strategies to improve speech perception in noise, most of them improve speech quality but do not improve speech intelligibility, in circumstances where the noise and speech have similar frequency spectra. Current exceptions that may improve speech intelligibility are those that require a priori knowledge of the speech or noise statistics, which limits practical application. Hearing impaired (HI) listeners suffer more in speech intelligibility than normal hearing listeners (NH) in the same noisy environment, so developing better single-channel noise reduction algorithms for HI listeners is justified. Our model-based "sparse coding shrinkage" (SCS) algorithm extracts key speech information in noisy speech. We evaluate it by comparison with a state-of-the-art Wiener filtering approach using speech intelligibility tests with NH and HI listeners. The model-based SCS algorithm relies only on statistical signal information without prior information. Results show that the SCS algorithm improves speech intelligibility in stationary noise and is comparable to the Wiener filtering algorithm. Both algorithms improve intelligibility for HI listeners but not for NH listeners. Improvement is less in fluctuating (babble) noise than in stationary noise. Both noise reduction algorithms perform better at higher input signal-to-noise ratios (SNR) where HI listeners can benefit but where NH listeners have already reached ceiling performance. The difference between NH and HI subjects in intelligibility gain depends fundamentally on the input SNR rather than the hearing loss level. We conclude that HI listeners need different signal processing algorithms from NH subjects and that the SCS algorithm offers a promising alternative to Wiener filtering. Performance of all noise reduction algorithms is likely to vary according to extent of hearing loss and algorithms that show little benefit for listeners with moderate hearing loss may be more beneficial for listeners with more severe hearing loss.

Improved Sparse Coding Under the Influence of Perceptual Attention

Sparse coding has established itself as a useful tool for the representation of natural data in the neuroscience as well as signal-processing literature. The aim of this letter, inspired by the human brain, is to improve on the performance of the sparse coding algorithm by trying to bridge the gap between neuroscience and engineering. To this end, we build on the localized perception-action cycle in cognitive neuroscience by categorizing it under the umbrella of perceptual attention, which lends itself to increase gradually the contrast between relevant information and irrelevant information. Stated in another way, irrelevant information is filtered away, while relevant information about the environment is enhanced from one cycle to the next. We may thus think in terms of the information filter, which, in a Bayesian context, was introduced in the literature by Fraser ( 1967 ). In a Bayesian context, the information filter provides a method for algorithmic implementation of perceptual attention. The information filter may therefore be viewed as the basis for improving the algorithmic performance of sparse coding. To support this performance improvement, the letter presents two computer experiments. The first experiment uses simulated (real-valued) data that are generated to purposely make the problem challenging. The second uses real-life radar data that are complex valued, hence the proposal to introduce Wirtinger calculus into derivation of the new algorithm.

A method for environmental acoustic analysis improvement based on individual evaluation of common sources in urban areas

Noise levels of common sources such as vehicles, whistles, sirens, car horns and crowd sounds are mixed in urban soundscapes. Nowadays, environmental acoustic analysis is performed based on mixture signals recorded by monitoring systems. These mixed signals make it difficult for individual analysis which is useful in taking actions to reduce and control environmental noise. This paper aims at separating, individually, the noise source from recorded mixtures in order to evaluate the noise level of each estimated source. A method based on blind deconvolution and blind source separation in the wavelet domain is proposed. This approach provides a basis to improve results obtained in monitoring and analysis of common noise sources in urban areas. The method validation is through experiments based on knowledge of the predominant noise sources in urban soundscapes. Actual recordings of common noise sources are used to acquire mixture signals using a microphone array in semi-controlled environments. The developed method has demonstrated great performance improvements in identification, analysis and evaluation of common urban sources.

Visible/near infrared spectroscopy and chemometrics for the prediction of trace element (Fe and Zn) levels in rice leaf

Two sensitive wavelength (SW) selection methods combined with visible/near infrared (Vis/NIR) spectroscopy were investigated to determine the levels of some trace elements (Fe, Zn) in rice leaf. A total of 90 samples were prepared for the calibration (n = 70) and validation (n = 20) sets. Calibration models using SWs selected by LVA and ICA were developed and nonlinear regression of a least squares-support vector machine (LS-SVM) was built. In the nonlinear models, six SWs selected by ICA can provide the optimal ICA-LS-SVM model when compared with LV-LS-SVM. The coefficients of determination (R²), root mean square error of prediction (RMSEP) and bias by ICA-LS-SVM were 0.6189, 20.6510 ppm and −12.1549 ppm, respectively, for Fe, and 0.6731, 5.5919 ppm and 1.5232 ppm, respectively, for Zn. The overall results indicated that ICA was a powerful way for the selection of SWs, and Vis/NIR spectroscopy combined with ICA-LS-SVM was very efficient in terms of accurate determination of trace elements in rice leaf.

Consciousness and depth of anesthesia assessment based on Bayesian analysis of EEG signals

This study applies Bayesian techniques to analyze EEG signals for the assessment of the consciousness and depth of anesthesia (DoA). This method takes the limiting large-sample normal distribution as posterior inferences to implement the Bayesian paradigm. The maximum a posterior (MAP) is applied to denoise the wavelet coefficients based on a shrinkage function. When the anesthesia states change from awake to light, moderate, and deep anesthesia, the MAP values increase gradually. Based on these changes, a new function B(DoA) is designed to assess the DoA. The new proposed method is evaluated using anesthetized EEG recordings and BIS data from 25 patients. The Bland-Alman plot is used to verify the agreement of B(DoA) and the popular BIS index. A correlation between B(DoA) and BIS was measured using prediction probability P(K). In order to estimate the accuracy of DoA, the effect of sample n and variance τ on the maximum posterior probability is studied. The results show that the new index accurately estimates the patient's hypnotic states. Compared with the BIS index in some cases, the B(DoA) index can estimate the patient's hypnotic state in the case of poor signal quality.

Framelet algorithms for de-blurring images corrupted by impulse plus Gaussian noise

This paper studies a problem of image restoration that observed images are contaminated by Gaussian and impulse noise. Existing methods for this problem in the literature are based on minimizing an objective functional having the l(1) fidelity term and the Mumford-Shah regularizer. We present an algorithm on this problem by minimizing a new objective functional. The proposed functional has a content-dependent fidelity term which assimilates the strength of fidelity terms measured by the l(1) and l(2) norms. The regularizer in the functional is formed by the l(1) norm of tight framelet coefficients of the underlying image. The selected tight framelet filters are able to extract geometric features of images. We then propose an iterative framelet-based approximation/sparsity deblurring algorithm (IFASDA) for the proposed functional. Parameters in IFASDA are adaptively varying at each iteration and are determined automatically. In this sense, IFASDA is a parameter-free algorithm. This advantage makes the algorithm more attractive and practical. The effectiveness of IFASDA is experimentally illustrated on problems of image deblurring with Gaussian and impulse noise. Improvements in both PSNR and visual quality of IFASDA over a typical existing method are demonstrated. In addition, Fast_IFASDA, an accelerated algorithm of IFASDA, is also developed.

A linear structural equation model for covert verb generation based on independent component analysis of FMRI data from children and adolescents

Human language is a complex and protean cognitive ability. Young children, following well defined developmental patterns learn language rapidly and effortlessly producing full sentences by the age of 3 years. However, the language circuitry continues to undergo significant neuroplastic changes extending well into teenage years. Evidence suggests that the developing brain adheres to two rudimentary principles of functional organization: functional integration and functional specialization. At a neurobiological level, this distinction can be identified with progressive specialization or focalization reflecting consolidation and synaptic reinforcement of a network (Lenneberg, 1967; Muller et al., 1998; Berl et al., 2006). In this paper, we used group independent component analysis and linear structural equation modeling (McIntosh and Gonzalez-Lima, 1994; Karunanayaka et al., 2007) to tease out the developmental trajectories of the language circuitry based on fMRI data from 336 children ages 5–18 years performing a blocked, covert verb generation task. The results are analyzed and presented in the framework of theoretical models for neurocognitive brain development. This study highlights the advantages of combining both modular and connectionist approaches to cognitive functions; from a methodological perspective, it demonstrates the feasibility of combining data-driven and hypothesis driven techniques to investigate the developmental shifts in the semantic network.

Using Watermarking Techniques to prove Rightful Ownership of Web Images

This article introduces intelligent watermarking scheme to protect Web images from attackers who try to counterfeit the copyright to damage the rightful ownership. Using secret signs and logos that are embedded within the digital images, the technique can investigate technically the ownership claim. Also, the nature of each individual image is taken into consideration which gives more reliable results. The colour channel used was chosen depending on the value of its standard deviation to compromise between robustness and invisibility of the watermarks. Several types of test images, logos, attacks and evaluation metrics were used to examine the performance of the techniques used. Subjective and objective tests were used to check visually and mathematically the solidity and weakness of the used scheme.

Iterative Gaussianization: from ICA to random rotations

Most signal processing problems involve the challenging task of multidimensional probability density function (PDF) estimation. In this paper, we propose a solution to this problem by using a family of rotation-based iterative Gaussianization (RBIG) transforms. The general framework consists of the sequential application of a univariate marginal Gaussianization transform followed by an orthonormal transform. The proposed procedure looks for differentiable transforms to a known PDF so that the unknown PDF can be estimated at any point of the original domain. In particular, we aim at a zero-mean unit-covariance Gaussian for convenience. RBIG is formally similar to classical iterative projection pursuit algorithms. However, we show that, unlike in PP methods, the particular class of rotations used has no special qualitative relevance in this context, since looking for interestingness is not a critical issue for PDF estimation. The key difference is that our approach focuses on the univariate part (marginal Gaussianization) of the problem rather than on the multivariate part (rotation). This difference implies that one may select the most convenient rotation suited to each practical application. The differentiability, invertibility, and convergence of RBIG are theoretically and experimentally analyzed. Relation to other methods, such as radial Gaussianization, one-class support vector domain description, and deep neural networks is also pointed out. The practical performance of RBIG is successfully illustrated in a number of multidimensional problems such as image synthesis, classification, denoising, and multi-information estimation.

Psychophysically tuned divisive normalization approximately factorizes the PDF of natural images

The conventional approach in computational neuroscience in favor of the efficient coding hypothesis goes from image statistics to perception. It has been argued that the behavior of the early stages of biological visual processing (e.g., spatial frequency analyzers and their nonlinearities) may be obtained from image samples and the efficient coding hypothesis using no psychophysical or physiological information. In this work we address the same issue in the opposite direction: from perception to image statistics. We show that psychophysically fitted image representation in V1 has appealing statistical properties, for example, approximate PDF factorization and substantial mutual information reduction, even though no statistical information is used to fit the V1 model. These results are complementary evidence in favor of the efficient coding hypothesis.

Principal neighborhood dictionaries for nonlocal means image denoising

We present an in-depth analysis of a variation of the nonlocal means (NLM) image denoising algorithm that uses principal component analysis (PCA) to achieve a higher accuracy while reducing computational load. Image neighborhood vectors are first projected onto a lower dimensional subspace using PCA. The dimensionality of this subspace is chosen automatically using parallel analysis. Consequently, neighborhood similarity weights for denoising are computed using distances in this subspace rather than the full space. The resulting algorithm is referred to as principal neighborhood dictionary (PND) nonlocal means. We investigate PND's accuracy as a function of the dimensionality of the projection subspace and demonstrate that denoising accuracy peaks at a relatively low number of dimensions. The accuracy of NLM and PND are also examined with respect to the choice of image neighborhood and search window sizes. Finally, we present a quantitative and qualitative comparison of PND versus NLM and another image neighborhood PCA-based state-of-the-art image denoising algorithm.

Statistical models of natural images and cortical visual representation

A fundamental question in visual neuroscience is: Why are the response properties of visual neurons as they are? A modern approach to this problem emphasizes the importance of adaptation to ecologically valid input, and it proceeds by modeling statistical regularities in ecologically valid visual input (natural images). A seminal model was linear sparse coding, which is equivalent to independent component analysis (ICA), and provided a very good description of the receptive fields of simple cells. Further models based on modeling residual dependencies of the ''independent" components have later been introduced. These models lead to emergence of further properties of visual neurons: the complex cell receptive fields, the spatial organization of the cells, and some surround suppression and Gestalt effects. So far, these models have concentrated on the response properties of neurons, but they hold great potential to model various forms of inference and learning.

Optimal Approximation of Signal Priors

In signal restoration by Bayesian inference, one typically uses a parametric model of the prior distribution of the signal. Here, we consider how the parameters of a prior model should be estimated from observations of uncorrupted signals. A lot of recent work has implicitly assumed that maximum likelihood estimation is the optimal estimation method. Our results imply that this is not the case. We first obtain an objective function that approximates the error occurred in signal restoration due to an imperfect prior model. Next, we show that in an important special case (small gaussian noise), the error is the same as the score-matching objective function, which was previously proposed as an alternative for likelihood based on purely computational considerations. Our analysis thus shows that score matching combines computational simplicity with statistical optimality in signal restoration, providing a viable alternative to maximum likelihood methods. We also show how the method leads to a new intuitive and geometric interpretation of structure inherent in probability distributions.

Optimal denoising in redundant representations

Image denoising methods are often designed to minimize mean-squared error (MSE) within the subbands of a multiscale decomposition. However, most high-quality denoising results have been obtained with overcomplete representations, for which minimization of MSE in the subband domain does not guarantee optimal MSE performance in the image domain. We prove that, despite this suboptimality, the expected image-domain MSE resulting from applying estimators to subbands that are made redundant through spatial replication of basis functions (e.g., cycle spinning) is always less than or equal to that resulting from applying the same estimators to the original nonredundant representation. In addition, we show that it is possible to further exploit overcompleteness by jointly optimizing the subband estimators for image-domain MSE. We develop an extended version of Stein's unbiased risk estimate (SURE) that allows us to perform this optimization adaptively, for each observed noisy image. We demonstrate this methodology using a new class of estimator formed from linear combinations of localized "bump" functions that are applied either pointwise or on local neighborhoods of subband coefficients. We show through simulations that the performance of these estimators applied to overcomplete subbands and optimized for image-domain MSE is substantially better than that obtained when they are optimized within each subband. This performance is, in turn, substantially better than that obtained when they are optimized for use on a nonredundant representation.

Dose detection of radiated rice by infrared spectroscopy and chemometrics

Infrared spectroscopy based on sensitive wavelengths (SWs) and chemometrics was proposed to discriminate the nine different radiation doses (0, 250, 500, 750, 1000, 1500, 2000, 2500, and 3000 Gy) of rice. Samples ( n = 16 each dose) were selected randomly for the calibration set, and the remaining 36 samples ( n = 4 each dose) were selected for the prediction set. Partial least-squares (PLS) analysis and least-squares-support vector machine (LS-SVM) were implemented for calibration models. PLS analysis was implemented for calibration models with different wavelength bands including near-infrared (NIR) regions and mid-infrared (MIR) regions. The best PLS models were achieved in the MIR (400-4000 cm (-1)) region. Furthermore, different latent variables (5-9 LVs) were used as inputs of LS-SVM to develop the LV-LS-SVM models with a grid search technique and radial basis function (RBF) kernel. The optimal models were achieved with six LVs, and they outperformed PLS models. Moreover, independent component analysis (ICA) was executed to select several SWs based on loading weights. The optimal LS-SVM model was achieved with SWs (756, 895, 1140, and 2980 cm (-1)) selected by ICA and had better performance than PLS and LV-LS-SVM with the parameters of correlation coefficient ( r), root-mean-square error of prediction, and bias of 0.996, 80.260, and 5.172 x 10 (-4), respectively. The overall results indicted that the ICA was an effective way for the selection of SWs, and infrared spectroscopy combined with LS-SVM models had the capability to predict the different radiation doses of rice.

Robustness of neural codes and its implication on natural image processing

In this study, based on the view of statistical inference, we investigate the robustness of neural codes, i.e., the sensitivity of neural responses to noise, and its implication on the construction of neural coding. We first identify the key factors that influence the sensitivity of neural responses, and find that the overlap between neural receptive fields plays a critical role. We then construct a robust coding scheme, which enforces the neural responses not only to encode external inputs well, but also to have small variability. Based on this scheme, we find that the optimal basis functions for encoding natural images resemble the receptive fields of simple cells in the striate cortex. We also apply this scheme to identify the important features in the representation of face images and Chinese characters.

Comparing methods to denoise mammographic images

Digital mammographic image processing often requires a previous application of filters to reduce the noise level of the image while preserving important details. This may improve the quality of digital mammographic images and contribute to an accurate diagnosis. In the literature, one can find a large amount of denoising techniques available for different kinds of images. We have adapted some of the existing denoising algorithms to mammographic images. We compare the effect of different denoising filters acting on digitized mammograms. The considered filters are: a local Wiener filter, a wavelet filter, a filter based on independent component analysis, and finally, a filter based on the diffusion equation. The noise reduction is measured by the mean squared error.