Analysis of mixed lipid extracts using 1H NMR spectra

A method has been developed for rapid analysis of lipid proton NMR spectra. Identification of lipid content is possible because of the presence of unique peaks or ratios of peaks for individual lipids. Spectra can be subdivided into regions where peaks represent certain chemical groups held in common, or uniquely by the various lipids. Vectors (B) are made up of the areas of these subdivisions of peaks from spectra of unknown components. A new FORTRAN algorithm, LIPICK, tests for the presence of unique peaks or combinations of peaks to determine which lipids may be present. The spectra vectors of known identified lipids are then placed in the (A) matrix of possible solution candidates. Quantitation of lipids in an 1H NMR spectrum (B) of an unknown mixture then proceeds by solving the equation AX = B for X (the concentrations of the individual lipids present) by singular value analysis. At this time, it is possible to test 1 mg of total lipid for the presence and relative concentration of 15 common lipids: cholesterol and its esters; phosphatidyl-ethanolamine, -choline, -serine, -inositol, -glycerol; tri- or di-glycerides; fatty acid; lysophosphatidylcholine; sphingomyelin; cerebrosides and sulfatides; dolichol and dolichol P; and phosphatidic acid. This procedure is suitable for membrane lipid analysis and has been evaluated using known mixtures of purified standard lipids.     

基于双特征高斯混合模型和双约束空间变换的配准

非刚性点集配准是当前多个领域中的一项重要研究问题.现今流行的配准算法通常使用基于单一特征的对应关系评估与包含单一约束条件的空间变换更新,而单特征与单约束限制了其配准效果与应用领域.提出了一种基于双特征高斯混合模型和双约束空间变换的非刚性点集配准算法.首先定义了双特征描述子,并用全局特征和局部特征构建它;随后,基于此描述子将高斯混合模型改进为双特征高斯混合模型.定义了局部结构约束项,并与全局结构约束项分别维护点集在进行空间变换更新时的局部与全局结构稳定.通过交替进行基于双特征高斯混合模型评估点集之间的对应关系和基于高斯径向基函数（Gaussian radial basis function）更新双约束空间变换,使该算法准确地完成非刚性点集配准.通过人造点集配准、CMU序列图像配准、遥感图像配准、IMM人脸数据配准和真实图像特征点配准对该算法进行了性能测试,同时也与当前流行的8种算法进行了性能比较实验,该算法展现出了卓越的非刚性配准性能,并在大部分实验中超越了当前的相关算法.     

Variance estimation and ranking of target tracking position errors modeled using Gaussian mixture distributions

In this paper, variance estimation and ranking methods are developed for stochastic processes modeled by Gaussian mixture distributions. It is shown that the variance estimate from a Gaussian mixture distribution has the same properties as the variance estimate from a single Gaussian distribution based on a reduced number of samples. Hence, well-known tools for variance estimation and ranking of single Gaussian distributions can be applied to Gaussian mixture distributions. As an application example, we present optimization of sensor processing order in the sequential multi-target multi-sensor joint probabilistic data association (MSJPDA) algorithm.     

Field spectroscopy of estuarine intertidal habitats

The recent introduction of portable, low‐cost hyperspectral radiometers for measuring the reflectance of marine intertidal habitats has considerable promise, first as a source of reference spectra for airborne and satellite remote sensing, and second as a survey technique in its own right. This paper reports on the results of an intercalibration exercise in which two designs of commercially available instruments were used to measure reflectance spectra across a diverse, intertidal site. The substrates sampled included seagrasses, macroalgae, and sediments which varied in wetness and grain size. Spectra were compared statistically using several measures of similarity, and the results showed a high degree of correlation between the measured reflectance spectra. There were no significant differences between instruments from different manufacturers in the shape of reflectance spectra, but the signal‐to‐noise ratio varied considerably between individual instruments. Spectra measured by operators with lesser experience in field spectrometry tended to be more variable than those of more experienced operators, indicating that training and adherence to set protocols are important when implementing this method in the field situation. Spatial plots derived from field spectra showed essentially the same trends in surface features as plots created using traditional sampling and remote sensing methods. Furthermore, the use of field spectrometry as a stand‐alone, low‐cost method for rapidly mapping the distribution of major habitat types was demonstrated by the results of a survey of a large intertidal sandbank.     

Reproducing Spectral Reflectances From Tristimulus Colours

Physically based rendering systems often support spectral rendering to simulate light transport in the real world. Material representations in such simulations need to be defined as spectral distributions. Since commonly available material data are in tristimulus colours, we ideally would like to obtain spectral distributions from tristimulus colours as an input to spectral rendering systems. Reproduction of spectral distributions given tristimulus colours, however, has been considered an ill‐posed problem since single tristimulus colour corresponds to a set of different spectra due to metamerism. We show how to resolve this problem using a data‐driven approach based on measured spectra and propose a practical algorithm that can faithfully reproduce a corresponding spectrum only from the given tristimulus colour. The key observation in colour science is that a natural measured spectrum is usually well approximated by a weighted sum of a few basis functions. We show how to reformulate conversion of tristimulus colours to spectra via principal component analysis. To improve accuracy of conversion, we propose a greedy clustering algorithm which minimizes reconstruction error. Using pre‐computation, the runtime computation is just a single matrix multiplication with an input tristimulus colour. Numerical experiments show that our method well reproduces the reference measured spectra using only the tristimulus colours as input.     

谱图解析中重叠峰分离的计算方法研究(Ⅰ)

本文通过对谱峰分离的拟合方法所涉及的数学规划问题的研究,提出M个重叠峰分离计算应作为有约束非线性规划问题看待;提出了DRAT(降维-绝对值变换)计算策略,它通过构造线性模型参数的极小化点集并对此进行绝对值变换,使得一个3M+1维有约束规划问题转化成一个等价的2M维无约束规划问题;在此基础上,应用改进Powell方法、BFGS变尺度方法以及二者的偶合方法,成功地在微机上实现了对一些典型的二重叠直至七重叠峰的完全分离。理论分析和实算结果表明,将谱峰分离计算处理成有约束规划问题,可以有效地避免由于峰参数初值估计不当所引起的假收敛;DRAT计算方法,由于考虑了约束的作用并且优化过程不直接涉及峰强参数,因而具有较强的收敛性和很宽的可行初值范围,从而它为各类谱图解析中的重叠峰分离工作提供了一种有效的计算方法。     

Technical Note: Equivalence of modified Gaussian model (MGM) in wavenumber and Gaussian in wavelength for deconvolution of hyperspectral reflectance spectra

A modified Gaussian model (MGM) is used to separate overlapping absorption features into their fundamental bands in mineral reflectance spectra; it is widely used in the planetary remote sensing community. To facilitate its use for earth‐observation studies where it has not been widely adopted, we show that the MGM analysis in wavenumber is numerically equivalent to Gaussian analysis in wavelength, improving access to software that is capable of the analysis. The two approaches were tested on a synthetic spectrum and a measured mineral spectrum. Results for the two methods are as close to identical as computationally possible, confirming their equivalency.     

复子波提取语音信号特征信息

对时变性强的非平稳汉语语音信号,建议采用变时-频复子波分析方法提取汉语语音信号的幅度谱、相位谱、基音周期及共振峰信息。选择有n阶消失矩及良好的时频局域化特性的复高斯子波提取汉语语音信号的幅度谱和相位谱,实验结果表明,该方法提取的语音信号的幅度谱、相位谱和子波变换谱表征了汉语语音的音节包络、细节包络及声调,区分了清、浊音,并准确提取了语音信号的动态基音周期、估计出共振峰。这对汉语语音特征提取和识别提供了一种新的思路。     

Performance comparison of multitaper techniques for speaker verification with expressive speech

In this paper, we provide a comparative study of spectral front-end features used as representations for speech signals by processing multitaper magnitude and phase spectra, for speaker verification with expressive speech. In particular, the multitaper modified group delay function (MT-MOGDF) and multitaper magnitude (MT-MAG) spectra of the speech signals are employed to obtain low variance estimates of speech spectra. We observe that the cues that aid in representation of expressive speech are evident in the MT-MOGDF spectrum than the MT-MAG spectrum in terms of mean Formant value and Formant bandwidth. Our extensive experimental study on a speaker verification system with a Gaussian mixture model based universal background model classifier on expressive speech using the IITKGP-SESC and EMODB databases show that MT-MOGDF performs better than MT-MAG technique, in terms of equal error rate and minimum decision cost function. This improvement due to MT-MOGDF is owed to a better representation and a low-variance estimate of the speech spectrum. Our results highlight the utility of MT-MOGDF as a potential alternative for MT-MAG representation for speaker verification problems in general.     

The identification of phytoplankton pigments from absorption spectra

The absorption spectra of photosynthetic algae are characterized by a continuous envelope, which is a result of the overlapping spectra of the indivual pigments. This feature makes it difficult to estimate the contribution of each pigment to the total absorption spectra. Derivative analysis is an objective tool for isolating absorption peaks in phytoplankton. Theoretically, electrons and ions of interacting molecules can be regarded as simple harmonic oscillators in an electromagnetic field, which result in a Lorentzian shape. However, when measured by an optical spectrophotometer the signal is transformed into a Gaussian curve. Thus, a combination of both types of curve provides a realistic approach to the decomposition of absorption spectra. In this study derivative analysis is performed on absorption spectra in order to prove that the method can be successfully used to identify the individual absorption spectra of component pigments. The spectra used are modelled phytoplankton, spectrophotometric measurements of algal cultures and samples from natural waters. A combination of Gaussian-Lorentzian shaped curves, centred on the identified peaks, were compared with the original spectra and showed good agreement.     

层次混合模型快速遥感影像分割算法

针对高分辨率遥感影像中同物异谱和同谱异物导致的分割困难问题,提出一种层次高斯混合模型(HGMM)快速遥感影像分割算法.首先采用HGMM构建影像的统计模型,其具有准确建模像素强度统计分布呈现的非对称、重尾和多峰等复杂特性的能力;然后根据贝叶斯理论构建基于HGMM的分割模型,为了简化参数求解并提高算法效率,定义均值和方差为关于权重的函数;最后采用共轭梯度(CGM)求解模型参数.实验中采用所提出算法和传统统计模型分割算法分别对合成、全色和彩色高分辨率遥感影像进行分割实验.实验结果表明,所提出的HGMM具有准确建模复杂统计分布的能力,且能够准确和有效地分割全色和彩色遥感影像.     

基于形态学的高斯模型和八邻域帧差法混合运动目标检测算法

针对视频中运动目标的提取问题,提出一种基于形态学的高斯模型和八邻域帧差法相融合的提取算法。该算法首先将视频中某些帧转化为灰度图,建立以混合高斯分布为基础的统计模型,并结合八邻域帧差法提取出运动目标的大致轮廓,然后利用自适应更新的高斯模型算法进行精确的减除,最后再进行形态学处理,从而使检测出的运动目标更加清晰完整。实验结果表明,该算法对含有低速运动物体、阴影较多的视频提取效果较好,具有很好的鲁棒性。     

Curve-fit analysis of chlorophyll fluorescence spectra: Application to nutrient stress detection in sunflower

The in vivo laser-induced chlorophyll fluorescence (LICF) spectra of healthy and nutrient-deficient sunflower plants were measured on a Jobin Yvon monochromator with He---Ne laser excitation. To correctly determine the peak center and to evaluate the relative contributions of the bands in the total fluorescence spectrum, the steady state LICF spectra were analyzed with a nonlinear iterative procedure using Gaussian, Lorentzian, Pearson, Voigt, and exponential Gaussian spectral functions. It was observed that curve fitting performed by using two Gaussian peaks centered at 690 and 730 nm usually fits well to the chlorophyll fluorescence spectra. After curve fitting, the mean peak centers of the red and far-red chlorophyll bands of control plants were observed at 688.2 and 725.4 nm, respectively. A blue shift of as much as 9 nm in the peak position of the far-red band was observed with nutrient stress, whereas the shift in position of the red band was only of the order of a few nanometers. Further, the width at half maximum of the far-red band was found to increase by as much as 20 nm with nutrient stress. Curve fitting could thus separate out the red and far-red fluorescence spectra from a pair of normally distributed curves centered at 690 and 730 nm wavelengths, thereby differentiating the effects due to reabsorption from those due directly to changes in photosynthetic electron transport. The F690/F730 fluorescence intensity ratio obtained from curve-fitted parameters was found to be more sensitive to plant stress than were fluorescence values alone. Results indicate that a curve-fitting analysis of LICF spectra using Gaussian spectral functions is a very useful and sensitive method of evaluating spectral features from a statistical point of view and for accurate determination of contributions from constituent bands in the whole leaf fluorescence spectrum.     

基于混合高斯模型MRF场的CT图像分割*

提出了一种基于混合高斯模型的马尔可夫随机场CT图像分割方法.此方法根据工业CT图像的特点,建立混合高斯逼近的图像灰度统计模型;用混合高斯模型作为Markov随机场的先验模型,提出混合高斯Markov随机场分割模型.实验表明,该方法较单高斯模型有很大的改善,对工业CT图像分割效果好.     

Using redundant parallel architecture to improve speaker recognition performance

In this paper, we propose two kinds of modifications in speaker recognition. First, the correlations between frequency channels are of prime importance for speaker recognition. Some of these correlations are lost when the frequency domain is divided into sub-bands. Consequently we propose a particularly redundant parallel architecture for which most of the correlations are kept. Second, generally a log transformation used to modify the power spectrum is done after the filter-bank in the classical spectrum calculation. We will see that performing this transformation before the filter bank is more interesting in our case. In the processing of recognition, the Gaussian mixture model (GMM) recognition arithmetic is adopted. Experiments on speech corrupted by noise show a better adaptability of this approach in noisy environments, compared with a conventional device, especially when pruning of some recognizers is performed.     

Using redundant parallel architecture to improve speaker recognition performance

In this paper, we propose two kinds of modifications in speaker recognition. First, the correlations between frequency channels are of prime importance for speaker recognition. Some of these correlations are lost when the frequency domain is divided into sub-bands. Consequently we propose a particularly redundant parallel architecture for which most of the correlations are kept. Second, generally a log transformation used to modify the power spectrum is done after the filter-bank in the classical spectrum calculation. We will see that performing this transformation before the filter bank is more interesting in our case. In the processing of recognition, the Gaussian mixture model （GMM） recognition arithmetic is adopted. Experiments on speech corrupted by noise show a better adaptability of this approach in noisy environments, comoared with a conventional device, esoeciallv when oruning of some recognizers is performed.     

Bayesian analysis of an inverse Gaussian correlated frailty model

In survival analysis frailty is often used to model heterogeneity between individuals or correlation within clusters. Typically frailty is taken to be a continuous random effect, yielding a continuous mixture distribution for survival times. A Bayesian analysis of a correlated frailty model is discussed in the context of inverse Gaussian frailty. An MCMC approach is adopted and the deviance information criterion is used to compare models. As an illustration of the approach a bivariate data set of corneal graft survival times is analysed.     

基于高斯混合模型的DCT域水印检测方法

基于音频DCT系数的统计特征,提出了一种水印检测方法.采用扩频的方法在DCT域嵌入水印, 分别对嵌入水印和未包含水印的音频信号在DCT域进行统计学习,得到对应的高斯混合模型(Gaussian mixture model, GMM). 接收端采用最大似然检测,判断是否嵌入水印并提取相应的水印信息. 仿真结果表明本文的水印检测算法对常见的信号攻击具有鲁棒性,与传统的相关检测法相比,检测可靠性更高.     

Voice Conversion Based on Maximum-Likelihood Estimation of Spectral Parameter Trajectory

In this paper, we describe a novel spectral conversion method for voice conversion (VC). A Gaussian mixture model (GMM) of the joint probability density of source and target features is employed for performing spectral conversion between speakers. The conventional method converts spectral parameters frame by frame based on the minimum mean square error. Although it is reasonably effective, the deterioration of speech quality is caused by some problems: 1) appropriate spectral movements are not always caused by the frame-based conversion process, and 2) the converted spectra are excessively smoothed by statistical modeling. In order to address those problems, we propose a conversion method based on the maximum-likelihood estimation of a spectral parameter trajectory. Not only static but also dynamic feature statistics are used for realizing the appropriate converted spectrum sequence. Moreover, the oversmoothing effect is alleviated by considering a global variance feature of the converted spectra. Experimental results indicate that the performance of VC can be dramatically improved by the proposed method in view of both speech quality and conversion accuracy for speaker individuality.