Robust visual tracking using structurally random projection and weighted least squares

Sparse representation based visual tracking approaches have attracted increasing interests in the community in recent years. The main idea is to linearly represent each target candidate using a set of target and trivial templates while imposing a sparsity constraint onto the representation coefficients. After we obtain the coefficients using L1-norm minimization methods, the candidate with the lowest error, when it is reconstructed using only the target templates and the associated coefficients, is considered as the tracking result. In spite of promising system performance widely reported, it is unclear if the performance of these trackers can be maximised. In addition, computational complexity caused by the dimensionality of the feature space limits these algorithms in real-time applications. In this paper, we propose a real-time visual tracking method based on structurally random projection and weighted least squares techniques. In particular, to enhance the discriminative capability of the tracker, we introduce background templates to the linear representation framework. To handle appearance variations over time, we relax the sparsity constraint using a weighed least squares (WLS) method to obtain the representation coefficients. To further reduce the computational complexity, structurally random projection is used to reduce the dimensionality of the feature space while preserving the pairwise distances between the data points in the feature space. Experimental results show that the proposed approach outperforms several state-of-the-art tracking methods.

Two-Stage Orthogonal Least Squares Methods for Neural Network Construction

A number of neural networks can be formulated as the linear-in-the-parameters models. Training such networks can be transformed to a model selection problem where a compact model is selected from all the candidates using subset selection algorithms. Forward selection methods are popular fast subset selection approaches. However, they may only produce suboptimal models and can be trapped into a local minimum. More recently, a two-stage fast recursive algorithm (TSFRA) combining forward selection and backward model refinement has been proposed to improve the compactness and generalization performance of the model. This paper proposes unified two-stage orthogonal least squares methods instead of the fast recursive-based methods. In contrast to the TSFRA, this paper derives a new simplified relationship between the forward and the backward stages to avoid repetitive computations using the inherent orthogonal properties of the least squares methods. Furthermore, a new term exchanging scheme for backward model refinement is introduced to reduce computational demand. Finally, given the error reduction ratio criterion, effective and efficient forward and backward subset selection procedures are proposed. Extensive examples are presented to demonstrate the improved model compactness constructed by the proposed technique in comparison with some popular methods.

Building Support Vector Machines in the Context of Regularised Least Squares

This paper formulates a linear kernel support vector machine (SVM) as a regularized least-squares (RLS) problem. By defining a set of indicator variables of the errors, the solution to the RLS problem is represented as an equation that relates the error vector to the indicator variables. Through partitioning the training set, the SVM weights and bias are expressed analytically using the support vectors. It is also shown how this approach naturally extends to Sums with nonlinear kernels whilst avoiding the need to make use of Lagrange multipliers and duality theory. A fast iterative solution algorithm based on Cholesky decomposition with permutation of the support vectors is suggested as a solution method. The properties of our SVM formulation are analyzed and compared with standard SVMs using a simple example that can be illustrated graphically. The correctness and behavior of our solution (merely derived in the primal context of RLS) is demonstrated using a set of public benchmarking problems for both linear and nonlinear SVMs.

Difference equation approach to two-thermocouple sensor characterisation in constant velocity flow environments

Thermocouples are one of the most popular devices for temperature measurement due to their robustness, ease of manufacture and installation, and low cost. However, when used in certain harsh environments, for example, in combustion systems and engine exhausts, large wire diameters are required, and consequently the measurement bandwidth is reduced. This article discusses a software compensation technique to address the loss of high frequency fluctuations based on measurements from two thermocouples. In particular, a difference equation sDEd approach is proposed and compared with existing methods both in simulation and on experimental test rig data with constant flow velocity. It is found that the DE algorithm, combined with the use of generalized total least squares for parameter identification, provides better performance in terms of time constant estimation without any a priori assumption on the time constant ratios of the thermocouples.

Exploiting a priori time constant ratio information in difference equation two-thermocouple sensor characterisation

The characterization of thermocouple sensors for temperature measurement in varying-flow environments is a challenging problem. Recently, the authors introduced novel difference-equation-based algorithms that allow in situ characterization of temperature measurement probes consisting of two-thermocouple sensors with differing time constants. In particular, a linear least squares (LS) lambda formulation of the characterization problem, which yields unbiased estimates when identified using generalized total LS, was introduced. These algorithms assume that time constants do not change during operation and are, therefore, appropriate for temperature measurement in homogenous constant-velocity liquid or gas flows. This paper develops an alternative ß-formulation of the characterization problem that has the major advantage of allowing exploitation of a priori knowledge of the ratio of the sensor time constants, thereby facilitating the implementation of computationally efficient algorithms that are less sensitive to measurement noise. A number of variants of the ß-formulation are developed, and appropriate unbiased estimators are identified. Monte Carlo simulation results are used to support the analysis.

Development and validation of an HPLC method for the determination of spironolactone and its metabolites in paediatric plasma samples

Abstract An HPLC method has been developed and validated for the determination of spironolactone, 7a-thiomethylspirolactone and canrenone in paediatric plasma samples. The method utilises 200 µl of plasma and sample preparation involves protein precipitation followed by Solid Phase Extraction (SPE). Determination of standard curves of peak height ratio (PHR) against concentration was performed by weighted least squares linear regression using a weighting factor of 1/concentration2. The developed method was found to be linear over concentration ranges of 30–1000 ng/ml for spironolactone and 25–1000 ng/ml for 7a-thiomethylspirolactone and canrenone. The lower limit of quantification for spironolactone, 7a-thiomethylspirolactone and canrenone were calculated as 28, 20 and 25 ng/ml, respectively. The method was shown to be applicable to the determination of spironolactone, 7a-thiomethylspirolactone and canrenone in paediatric plasma samples and also plasma from healthy human volunteers.

Risk adjustment for hospital use using social security data: cross sectional small area analysis

Objectives: To identify demographic and socioeconomic determinants of need for acute hospital treatment at small area level. To establish whether there is a relation between poverty and use of inpatient services. To devise a risk adjustment formula for distributing public funds for hospital services using, as far as possible, variables that can be updated between censuses. Design: Cross sectional analysis. Spatial interactive modelling was used to quantify the proximity of the population to health service facilities. Two stage weighted least squares regression was used to model use against supply of hospital and community services and a wide range of potential needs drivers including health, socioeconomic census variables, uptake of income support and family credit, and religious denomination. Setting: Northern Ireland. Main outcome measure: Intensity of use of inpatient services. Results: After endogeneity of supply and use was taken into account, a statistical model was produced that predicted use based on five variables: income support, family credit, elderly people living alone, all ages standardised mortality ratio, and low birth weight. The main effect of the formula produced is to move resources from urban to rural areas. Conclusions: This work has produced a population risk adjustment formula for acute hospital treatment in which four of the five variables can be updated annually rather than relying on census derived data. Inclusion of the social security data makes a substantial difference to the model and to the results produced by the formula.

Soil Carbon stocks and schanges in the Republic of Ireland

The soil carbon (C) stock of the Republic of Ireland is estimated to have been 2048 Mt in 1990 and 2021 Mt in 2000. Peat holds around 53% of the soil C stock, but on 17% of the land area. The C density of soils (t C ha-1) is mapped at 2 km*2 km resolution. The greatest soil C densities occur where deep raised bogs are the dominant soil; in these grid squares C density can reach 3000 t C ha-1. Most of the loss of soil C between 1990 and 2000-up to 23 Mt C (1% of 1990 soil C stock)-was through industrial peat extraction. The average annual change in soil C stocks from 1990 to 2000 due to land use change was estimated at around 0.02% of the 1990 stock. Considering uncertainties in the data used to calculate soil C stocks and changes, the small average annual 'loss' could be regarded as 'no change'.

A Fast Nonlinear Model Identification Method

The identification of nonlinear dynamic systems using linear-in-the-parameters models is studied. A fast recursive algorithm (FRA) is proposed to select both the model structure and to estimate the model parameters. Unlike orthogonal least squares (OLS) method, FRA solves the least-squares problem recursively over the model order without requiring matrix decomposition. The computational complexity of both algorithms is analyzed, along with their numerical stability. The new method is shown to require much less computational effort and is also numerically more stable than OLS.

A fast hybrid algorithm for exoplanetary transit searches

We present a fast and efficient hybrid algorithm for selecting exoplanetary candidates from wide-field transit surveys. Our method is based on the widely used SysRem and Box Least-Squares (BLS) algorithms. Patterns of systematic error that are common to all stars on the frame are mapped and eliminated using the SysRem algorithm. The remaining systematic errors caused by spatially localized flat-fielding and other errors are quantified using a boxcar-smoothing method. We show that the dimensions of the search-parameter space can be reduced greatly by carrying out an initial BLS search on a coarse grid of reduced dimensions, followed by Newton-Raphson refinement of the transit parameters in the vicinity of the most significant solutions. We illustrate the method's operation by applying it to data from one field of the SuperWASP survey, comprising 2300 observations of 7840 stars brighter than V = 13.0. We identify 11 likely transit candidates. We reject stars that exhibit significant ellipsoidal variations caused indicative of a stellar-mass companion. We use colours and proper motions from the Two Micron All Sky Survey and USNO-B1.0 surveys to estimate the stellar parameters and the companion radius. We find that two stars showing unambiguous transit signals pass all these tests, and so qualify for detailed high-resolution spectroscopic follow-up.

Experimental characterization of the nonlinear optical and magneto-optical properties of interfaces

An effective ellipsometric technique to determine parameters that characterize second-harmonic optical and magneto-optical effects in centrosymmetric media within the electric-dipole approximation is proposed and outlined in detail. The parameters, which are ratios of components of the nonlinear-surface-susceptibility tensors, are obtained from experimental data related to the state of polarization of the second-harmonic-generated radiation as a function of the angle between the plane of incidence and the polarization plane of the incident, linearly polarized, fundamental radiation. Experimental details of the technique are described. A corresponding theoretical model is given as an example for a single isotropic surface assuming polycrystalline samples. The surfaces of air-Au and air-Ni (in magnetized and demagnetized states) have been investigated ex situ in ambient air, and the results are presented. A nonlinear, least-squares-minimization fitting procedure between experimental data and theoretical formulas has been shown to yield realistic, unambiguous results for the ratios corresponding to each of the above materials. Independent methods for verifying the validity of the fitting parameters are also presented. The influence of temporal variations at the surfaces on the state of polarization (due to adsorption, contamination, or oxidation) is also illustrated for the demagnetized air-Ni surface. (C) 2005 Optical Society of America

New Zealand chironomids as proxies for human-induced and natural environmental change: Transfer functionsfor temperature and lake production (chlorophyll a)

The analysis of chironomid taxa and environmental datasets from 46 New Zealand lakes identified temperature (February mean air temperature) and lake production (chlorophyll a (Chl a)) as the main drivers of chironomid distribution. Temperature was the strongest driver of chironomid distribution and consequently produced the most robust inference models. We present two possible temperature transfer functions from this dataset. The most robust model (weighted averaging-partial least squares (WA-PLS), n = 36) was based on a dataset with the most productive (Chl a > 10 lg l)1) lakes removed. This model produced a coefficient of determination (r2 jack) of 0.77, and a root mean squared error of prediction (RMSEPjack) of 1.31C. The Chl a transfer function (partial least squares (PLS), n = 37) was far less reliable, with an r2 jack of 0.49 and an RMSEPjack of 0.46 Log10lg l)1. Both of these transfer functions could be improved by a revision of the taxonomy for the New Zealand chironomid taxa, particularly the genus Chironomus. The Chironomus morphotype was common in high altitude, cool, oligotrophic lakes and lowland, warm, eutrophic lakes. This could reflect the widespread distribution of one eurythermic species, or the collective distribution of a number of different Chironomus species with more limited tolerances. The Chl a transfer function could also be improved by inputting mean Chl a values into the inference model rather than the spot measurements that were available for this study.

Prediction of adipose tissue composition using Raman spectroscopy: average properties and individual fatty acids.

Abstract: Raman spectroscopy has been used for the first time to predict the FA composition of unextracted adipose tissue of pork, beef, lamb, and chicken. It was found that the bulk unsaturation parameters could be predicted successfully [R-2 = 0.97, root mean square error of prediction (RMSEP) = 4.6% of 4 sigma], with cis unsaturation, which accounted for the majority of the unsaturation, giving similar correlations. The combined abundance of all measured PUFA (>= 2 double bonds per chain) was also well predicted with R-2 = 0.97 and RMSEP = 4.0% of 4 sigma. Trans unsaturation was not as well modeled (R-2 = 0.52, RMSEP = 18% of 4 sigma); this reduced prediction ability can be attributed to the low levels of trans FA found in adipose tissue (0.035 times the cis unsaturation level). For the individual FA, the average partial least squares (PLS) regression coefficient of the 18 most abundant FA (relative abundances ranging from 0.1 to 38.6% of the total FA content) was R-2 = 0.73; the average RMSEP = 11.9% of 4 sigma. Regression coefficients and prediction errors for the five most abundant FA were all better than the average value (in some cases as low as RMSEP = 4.7% of 4 sigma). Cross-correlation between the abundances of the minor FA and more abundant acids could be determined by principal component analysis methods, and the resulting groups of correlated compounds were also well-predicted using PLS. The accuracy of the prediction of individual FA was at least as good as other spectroscopic methods, and the extremely straightforward sampling method meant that very rapid analysis of samples at ambient temperature was easily achieved. This work shows that Raman profiling of hundreds of samples per day is easily achievable with an automated sampling system.

Multivariate prediction of clarified butter composition using Raman spectroscopy

Raman spectroscopy has been used to predict the abundance of the FA in clarified butterfat that was obtained from dairy cows fed a range of levels of rapeseed oil in their diet. Partial least squares regression of the Raman spectra against FA compositions obtained by GC showed good prediction for the five major (abundance >5%) FA with R-2=0.74-0.92 and a root mean SE of prediction (RMSEP) that was 5-7% of the mean. In general, the prediction accuracy fell with decreasing abundance in the sample, but the RMSEP was 1.25%. The Raman method has the best prediction ability for unsaturated FA (R-2=0.85-0.92), and in particular trans unsaturated FA (best-predicted FA was 18:1 tDelta9). This enhancement was attributed to the isolation of the unsaturated modes from the saturated modes and the significantly higher spectral response of unsaturated bonds compared with saturated bonds. Raman spectra of the melted butter samples could also be used to predict bulk parameters calculated from standard analyzes, such as iodine value (R-2=0.80) and solid fat content at low temperature (R-2=0.87). For solid fat contents determined at higher temperatures, the prediction ability was significantly reduced (R-2=0.42), and this decrease in performance was attributed to the smaller range of values in solid fat content at the higher temperatures. Finally, although the prediction errors for the abundances of each of the FA in a given sample are much larger with Raman than with full GC analysis, the accuracy is acceptably high for quality control applications. This, combined with the fact that Raman spectra can be obtained with no sample preparation and with 60-s data collection times, means that high-throughput, on-line Raman analysis of butter samples should be possible.