Abstract: Skin color based tracking techniques often assume a
static skin color model obtained either from an offline set of library
images or the first few frames of a video stream. These models
can show a weak performance in presence of changing lighting or
imaging conditions. We propose an adaptive skin color model based
on the Gaussian mixture model to handle the changing conditions.
Initial estimation of the number and weights of skin color clusters
are obtained using a modified form of the general Expectation
maximization algorithm, The model adapts to changes in imaging
conditions and refines the model parameters dynamically using spatial
and temporal constraints. Experimental results show that the method
can be used in effectively tracking of hand and face regions.
Abstract: The mitigation of crop loss due to damaging freezes requires accurate air temperature prediction models. An improved model for temperature prediction in Georgia was developed by including information on seasonality and modifying parameters of an existing artificial neural network model. Alternative models were compared by instantiating and training multiple networks for each model. The inclusion of up to 24 hours of prior weather information and inputs reflecting the day of year were among improvements that reduced average four-hour prediction error by 0.18°C compared to the prior model. Results strongly suggest model developers should instantiate and train multiple networks with different initial weights to establish appropriate model parameters.
Abstract: The purpose of this study is to derive parameters
estimating for the Lyman–Kutcher–Burman (LKB) normal tissue
complication probability (NTCP) model using analysis of scintigraphy
assessments and quality of life (QoL) measurement questionnaires for
the parotid gland (xerostomia). In total, 31 patients with
head-and-neck (HN) cancer were enrolled. Salivary excretion factor
(SEF) and EORTC QLQ-H&N35 questionnaires datasets are used for
the NTCP modeling to describe the incidence of grade 4 xerostomia.
Assuming that n= 1, NTCP fitted parameters are given as TD50= 43.6
Gy, m= 0.18 in SEF analysis, and as TD50= 44.1 Gy, m= 0.11 in QoL
measurements, respectively. SEF and QoL datasets can validate the
Quantitative Analyses of Normal Tissue Effects in the Clinic
(QUANTEC) guidelines well, resulting in NPV-s of 100% for the both
datasets and suggests that the QUANTEC 25/20Gy gland-spared
guidelines are suitable for clinical used for the HN cohort to
effectively avoid xerostomia.
Abstract: Time varying network induced delays in networked
control systems (NCS) are known for degrading control system-s
quality of performance (QoP) and causing stability problems. In
literature, a control method employing modeling of communication
delays as probability distribution, proves to be a better method. This
paper focuses on modeling of network induced delays as probability
distribution.
CAN and MIL-STD-1553B are extensively used to carry periodic
control and monitoring data in networked control systems.
In literature, methods to estimate only the worst-case delays for
these networks are available. In this paper probabilistic network
delay model for CAN and MIL-STD-1553B networks are given.
A systematic method to estimate values to model parameters from
network parameters is given. A method to predict network delay in
next cycle based on the present network delay is presented. Effect of
active network redundancy and redundancy at node level on network
delay and system response-time is also analyzed.
Abstract: Estimation of stormwater pollutants is a pre-requisite
for the protection and improvement of the aquatic environment and
for appropriate management options. The usual practice for the
stormwater quality prediction is performed through water quality
modeling. However, the accuracy of the prediction by the models
depends on the proper estimation of model parameters. This paper
presents the estimation of model parameters for a catchment water
quality model developed for the continuous simulation of stormwater
pollutants from a catchment to the catchment outlet. The model is
capable of simulating the accumulation and transportation of the
stormwater pollutants; suspended solids (SS), total nitrogen (TN) and
total phosphorus (TP) from a particular catchment. Rainfall and water
quality data were collected for the Hotham Creek Catchment (HTCC),
Gold Coast, Australia. Runoff calculations from the developed model
were compared with the calculated discharges from the widely used
hydrological models, WBNM and DRAINS. Based on the measured
water quality data, model water quality parameters were calibrated
for the above-mentioned catchment. The calibrated parameters are
expected to be helpful for the best management practices (BMPs)
of the region. Sensitivity analyses of the estimated parameters were
performed to assess the impacts of the model parameters on overall
model estimations of runoff water quality.
Abstract: Autoregressive Moving average (ARMA) is a parametric based method of signal representation. It is suitable for problems in which the signal can be modeled by explicit known source functions with a few adjustable parameters. Various methods have been suggested for the coefficients determination among which are Prony, Pade, Autocorrelation, Covariance and most recently, the use of Artificial Neural Network technique. In this paper, the method of using Artificial Neural network (ANN) technique is compared with some known and widely acceptable techniques. The comparisons is entirely based on the value of the coefficients obtained. Result obtained shows that the use of ANN also gives accurate in computing the coefficients of an ARMA system.
Abstract: This study discusses the effect of uncertainty on
production levels of a petrochemical complex. Uncertainly or
variations in some model parameters, such as prices, supply and
demand of materials, can affect the optimality or the efficiency of any
chemical process. For any petrochemical complex with many plants,
there are many sources of uncertainty and frequent variations which
require more attention. Many optimization approaches are proposed
in the literature to incorporate uncertainty within the model in order
to obtain a robust solution. In this work, a stability analysis approach
is applied to a deterministic LP model of a petrochemical complex
consists of ten plants to investigate the effect of such variations on
the obtained optimal production levels. The proposed approach can
determinate the allowable variation ranges of some parameters,
mainly objective or RHS coefficients, before the system lose its
optimality. Parameters with relatively narrow range of variations, i.e.
stability limits, are classified as sensitive parameters or constraints
that need accurate estimate or intensive monitoring. These stability
limits offer easy-to-use information to the decision maker and help in
understanding the interaction between some model parameters and
deciding when the system need to be re-optimize. The study shows
that maximum production of ethylene and the prices of intermediate
products are the most sensitive factors that affect the stability of the
optimum solution
Abstract: Most neural network (NN) models of human category learning use a gradient-based learning method, which assumes that locally-optimal changes are made to model parameters on each learning trial. This method tends to under predict variability in individual-level cognitive processes. In addition many recent models of human category learning have been criticized for not being able to replicate rapid changes in categorization accuracy and attention processes observed in empirical studies. In this paper we introduce stochastic learning algorithms for NN models of human category learning and show that use of the algorithms can result in (a) rapid changes in accuracy and attention allocation, and (b) different learning trajectories and more realistic variability at the individual-level.
Abstract: This article presents a detailed analysis and comparative
performance evaluation of model reference adaptive control systems.
In contrast to classical control theory, adaptive control methods allow
to deal with time-variant processes. Inspired by the works [1] and
[2], two methods based on the MIT rule and Lyapunov rule are
applied to a linear first order system. The system is simulated and
it is investigated how changes to the adaptation gain affect the
system performance. Furthermore, variations in the reference model
parameters, that is changing the desired closed-loop behaviour are
examinded.
Abstract: This work presents a fusion of Log Gabor Wavelet
(LGW) and Maximum a Posteriori (MAP) estimator as a speech
enhancement tool for acoustical background noise reduction. The
probability density function (pdf) of the speech spectral amplitude is
approximated by a Generalized Laplacian Distribution (GLD).
Compared to earlier estimators the proposed method estimates the
underlying statistical model more accurately by appropriately
choosing the model parameters of GLD. Experimental results show
that the proposed estimator yields a higher improvement in
Segmental Signal-to-Noise Ratio (S-SNR) and lower Log-Spectral
Distortion (LSD) in two different noisy environments compared to
other estimators.
Abstract: Intravitreal injection (IVI) is the most common treatment for eye posterior segment diseases such as endopthalmitis, retinitis, age-related macular degeneration, diabetic retinopathy, uveitis, and retinal detachment. Most of the drugs used to treat vitreoretinal diseases, have a narrow concentration range in which they are effective, and may be toxic at higher concentrations. Therefore, it is critical to know the drug distribution within the eye following intravitreal injection. Having knowledge of drug distribution, ophthalmologists can decide on drug injection frequency while minimizing damage to tissues. The goal of this study was to develop a computer model to predict intraocular concentrations and pharmacokinetics of intravitreally injected drugs. A finite volume model was created to predict distribution of two drugs with different physiochemical properties in the rabbit eye. The model parameters were obtained from literature review. To validate this numeric model, the in vivo data of spatial concentration profile from the lens to the retina were compared with the numeric data. The difference was less than 5% between the numerical and experimental data. This validation provides strong support for the numerical methodology and associated assumptions of the current study.