Cerebrovascular diseases such as stroke are among the most common causes of death and disability worldwide preventable treatable. Early detection strokes their rapid intervention play an important role in reducing burden disease improving clinical outcomes. In recent years, machine learning methods have attracted a lot attention they can be used to detect strokes. The aim this study is identify reliable methods, algorithms, features that help medical professionals make informed decisions...

Lung cancer is one of the most common causes deaths in modern world. Screening lung nodules essential for early recognition to facilitate treatment that improves rate patient rehabilitation. An increase accuracy during detection vital sustaining persistence, even though several research works have been conducted this domain. Moreover, classical system fails segment cells different sizes accurately and with excellent reliability. This paper proposes a sooty tern optimization algorithm-based...

Abstract Gas hydrates are progressively becoming a key concern when determining the economics of reservoir due to flow interruptions, as offshore reserves produced in ever deeper and colder waters. The creation hydrate plug poses equipment safety risks. No current existing models have feature accurately predicting kinetics gas multiphase system is encountered. In this work, Artificial Neural Networks (ANN) developed study predict effect on formation. Primarily, pure containing crude oil used...

In this study, we propose a specimen tube prototype and smart transport box using radio frequency identification (RFID) narrow band–Internet of Things (NB-IoT) technology to use in the Department Laboratory Medicine, King Chulalongkorn Memorial Hospital. Our proposed method replaces existing system, based on barcode technology, with shortage usage low reliability. addition, tube-tagged has not eliminated lost or incorrect delivery issues many laboratories. solution, passive RFID tag is...

A collaborative painting robot that can be used as an alternative to workers has been developed using a digital twin framework and its performance was demonstrated experimentally. The of the automatic simulates entire process estimates paint result before real execution. An operator view simulated with option either confirm or cancel task. If task is accepted, generates all parameters, including end effector trajectory robot, material flow spray mechanism. This ability means practiced in...

In this paper, we proposed a new two-dimensional linear discriminant analysis (2DLDA) method. Based on principle component (2DPCA), face image matrices do not need to be previously transformed into vector. way, the spatial information can preserved. Moreover, 2DLDA also allows avoiding small sample size (SSS) problem, thus overcoming traditional LDA. We combine 2DPCA and our of vectors framework. Our framework consists two steps: first project an input family projected via 2DPCA-based...

We propose a method for online Thai handwritten character recognition using HMMs and SVMs with score-space kernels. Score-space kernels are generalized Fisher based on underlying generative models, such as Gaussian mixture models (GMMs), which output distributions of each state in HMMs. Our system combines the advantages both discriminative classifiers. In first phase, used multi-classification, then applied to resolve any uncertainty remaining after first-pass HMM-based recognizer, but they...

At first, this paper is concerned with wavelet-based image denoising using Bayesian technique. In conventional process, the parameters of probability density function (PDF) are usually calculated from first few moments, mean and variance. part our work, a new algorithm based on Pearson Type VII random vectors proposed. This PDF used because it allows higher-order moments to be incorporated into noiseless wavelet coefficients' probabilistic model. One cruxes algorithms estimate variance clean...

In this paper, we proposed a new Two-Dimensional Linear Discriminant Analysis (2DLDA) method, based on Principle Component (2DPCA) concept. particular, 2D face image matrices do not need to be previously transformed into vector. way, the spatial information can preserved. Moreover, 2DLDA also allows avoiding Small Sample Size (SSS) problem, thus overcoming traditional LDA. We combine 2DPCA and our of principle component vectors framework. Our framework consists two steps: first project an...

This paper presents a new on-line recognition of Thai handwritten characters. Active researches in character are converged into two distinct methods, HMM and fuzzy logic classifier. The former showed poor rate due to shortcoming the latter is on difficulties establishing set rules cover whole handwriting styles. Our method proposed exploit better worlds (HMM distinctive feature based classifier). experimental result was shown an average improved from 89.1%(using HMM) 91.2 using our method.

In this paper, we proposed a class-specific subspace-based two-dimensional principal component analysis (2DPCA) for face recognition. 2DPCA, 2D image matrices do not need to be previously transformed into vector. way, the spatial information can preserved. Moreover, 2DPCA achieve higher performance than PCA both in recognition and representation task. However, are unsupervised techniques, no of class labels considered. Therefore, directions that maximize scatter data might as adequate...

In this paper, we proposed a novel technique for face recognition using Two-Dimensional Random Subspace Analysis (2DRSA), based on the Principal Component (2DPCA) and Method (RSM). conventional 2DPCA, image covariance matrix is directly calculated via 21) images in form, by concept of random variable. However, 2DPCA reduces dimension original only one directions, nor mally row direction. Thus, it needs many more coefficients representation than PCA. For solving problem, methods were...

This paper presents a software development tool which is capable of simulating the vertical transportation systems within buildings using elevators. The simulation composed passenger arrival model for single or group elevators, their control system, and graphical user interface (GUI). developed GUI can display statistical information passengers traffic pattern also animation elevator cars with number inside. current version allows both individual batch Poisson processes could represent...

In this paper, we perform comparison techniques for environmental sound classification with multilayer perceptron (MLP) and support vector machine (SVM), deep learning using new platforms, i.e., Scikit-learn Tensorflow, respectively. For feature-based classification, principal component analysis of short-time Fourier transform is used as our feature the front end to MLP SVM. learning-based convolution+pooling layers acting extractor from input image, while fully connected layer will act a...

The annual degradation rate (DR) of photovoltaics (PV) system is a critical factor to evaluate the energy performance and levelized cost electricity (LCOE) during its operation lifetime. However, DR particular strongly depends on technical configuration such as PV module array, inverter configuration, also climatic conditions. Therefore, real dataset necessary engineer in order estimate LCOE for system. This article presents group systems Bangkok, Thailand which share same monocrystalline...

In this paper, we perform comparison techniques for environmental sound classification with multilayer perceptron (MLP) and support vector machine (SVM), deep learning using new platforms, i.e., Scikit-Iearn Tensorflow, respectively. For feature-based classification, principal component analysis of short-time Fourier transform is used as our feature the front end to MLP SVM. learning-based convolution+pooling layers acting extractor from input image, while fully connected layer will act a...

Presents an on-line bilingual handwritten character recognition to classify both Thai and English characters using distinctive feature extraction. In this paper, we analytically derive features for Thai-English language classification. Decision tree diagram based on derived is then used in practical applications. addition, a classifier has been as the front-end recognizer order improve performance of terms complexity reduction accuracy. From experimental results, our implemented system can...

This paper presents image-denoising methods performed within wavelet domain scheme by incorporating neighboring coefficients, namely NeighShrink (G.Y. Chen et al., 2004), and at the same time, denoising image with bivariate shrinkage function. The idea of function (BiShrink (L. Sendur I.W. Selesnick, 2002)) is to model signal based on MAP estimation approach. In fact, can also be bivariately modeled MMSE estimator. first method this work incorporate BiShrink model, called here MMSE_BiShrink....

In this work, we present new Bayesian estimator for circularly-contoured Two-Sided Gamma random vector in additive white Gaussian noise (AWGN). This PDF is used view of the fact that it more peaked and tails are heavier to be incorporated probabilistic modeling wavelet coefficients. One cruxes image denoising methods estimate statistical parameters a shrinkage function. We employ maximum posterior (MAP) estimation calculate local variances with Rayleigh density prior observed distribution...

Age‐related macular degeneration (AMD) diagnosis using fundus images is one of the critical missions eye‐care screening program in many countries. Various proposed deep learning models have been studied for this research interest, which aim to achieve mission and outperform human‐based approaches. However, efforts are still required improvement model classification accuracy, sensitivity, specificity values. In study, we named as ViT‐AMD, based on latest Vision Transformer (ViT) structure, a...

In this paper, we proposed a novel technique for face recognition using image cross-covariance analysis (ICCA), based on the two-dimensional principal component (2DPCA) technique. conventional 2DPCA, covariance matrix is directly calculated via 2D images in form, by concept of random variable. We found that it not optimal solution 2DPCA framework. Because some useful information classification neglected. Thus, introduced an which generalized form matrix. This defined two variables. The first...