The combination of Raman spectroscopy and deep learning technology provides an automatic, rapid, accurate scheme for the clinical diagnosis pathogenic bacteria. However, accuracy existing methods is still limited because single fixed scales neural networks. We propose a network that can learn multi-scale features spectra by using automatic multi-receptive fields convolutional layers. This model based on expert knowledge discrimination information composed spectral peaks. enhance...

In most optical fiber distributed acoustic sensing (DAS) systems, to obtain the desired outcome, signal acquired by DAS systems normally needs be denoised. some applications, such as identifiation of position fast-moving targets, we need respond with sufficient speed. However, classical denoising algorithms do not work if signals are insufficiently collected within a short period (called short-time signal). To ideal results time window, propose an attention-based convolutional neural network...

Deep learning anomaly detection is important in distributed optical fiber acoustic sensing (DAS). However, more challenging than traditional tasks, due to the scarcity of true-positive data and vast imbalance irregularity within datasets. Furthermore, it impossible catalog all types anomalies, therefore, direct application supervised deficient. To overcome these problems, an unsupervised deep method that only learns normal features from ordinary events proposed. First, a convolutional...

Hyperspectral remote sensing imagery provides valuable and rich information to distinguish the characteristics of materials. However, this advantage hyperspectral often encounters problem a limited amount training samples, which is caused by difficulty manually labeling. Fortunately, spatial distribution surface objects can be integrated with spectral signature improve discriminative ability. In paper, 3-D Gaussian-Gabor feature extraction selection framework has been proposed for image...

Introduction In recent years there has been an increasing interest in the field of educational computer games (ECGs). Although ECGs have researched, more analyses still need to be performed on these variables check their effects language learning. Methods To this end, 74-third grade female state high school students from two schools Zanjan were selected through a two-stage cluster random sampling method. The number each class was 37. One classes (control group) trained traditionally and...

In this paper, a conventional Strapdown Inertial Navigation System (SINS) alignment method on disturbed base is analysed. A novel with an attitude tracking idea proposed for the rocking alignment. It considered in that algorithm should track real changes process, but not excessively restrain disturbance. According to idea, rapid devised base. algorithm, coarse carried out within 30 s inertial frame precision less than 2°, which meets Kalman filter linearization conditions well. Then ten...

Low altitude small UAV recently becomes a hot technology with rapid development and comprehensive application which raising many risks in Low-altitude airspace. Considering the characteristics of low UAV, previous approaches like radar radio are insufficient, this paper is devoted to detection by using images from low-cost cameras. Method detecting based on YOLO model two neural networks, ResNet DenseNet, designed performed. One dataset established accordingly. Experiment shows that method...

The residue from desulfurization and denitrification of exhaust gas treatment process with pyrolusite ore as absorbent is regarded a potential source iron manganese. In this study, an extraction proposed for recovery manganese ammonium sulphate roasting followed by sulphuric acid leaching. Firstly, the conversion mechanism was analyzed through mineral phase analysis products at different temperature means X-ray diffraction (XRD) technology. Then, parameters procedure such time, dosage,...

Abstract Background Antifreeze proteins (AFPs) are a group of that inhibit body fluids from growing to ice crystals and thus improve biological antifreeze ability. It is vital the survival living organisms in extremely cold environments. However, little research performed on sequences feature extraction selection for classification structure function prediction, which great significance. Results In this paper, predict proteins, representation weighted generalized dipeptide composition...

A sensitive method for simultaneous determination of dithionate and sulfate using ion chromatography equipped with a suppressed conductivity detector was developed. The flow rate 1.2 mL min–1, the mobile phase contained KOH, which gradient-generated by an automatic generator. detection limits were 0.63 0.42 mg L–1 sulfate, respectively. correlation coefficients calibration curves greater than 0.997. linearity ranges 5–200 L–1, accuracies in range 99.5–111%. proposed rapid, accurate, fully...

Due to the depletion of high-grade manganese ore, much attention has turned recover from low-grade ores; however, these are usually associated with high-iron content. Moreover, conventional process extracting Mn iron-rich low grade oxide ore focused on recovery Mn, while ignoring leaching characteristics Fe. Hence, simultaneous investigations behavior and separation efficiency both iron were carried out for attaining high content pyrolusite using simulated flue gas as reductant. The effects...

Local binary patterns (LBP) features extracted from hyperspectral imagery (HSI) have gained impressive performance in classification tasks, for which LBP got considerable attention. However, existing LBP-based methods utilized two-dimensional (2DLBP) that could capture gray variation signal space, did not excavate the contextual information hides spectral-spatial structure considering characterizes by three dimension. Aimed at this problem, paper presents a three-dimensional (3DLBP) method...

Human head pose estimation is a hot topic in computer vision field, which can be used video surveillance, Computer Interaction and so on. Active Shape Model template matching method, suitable for object localization point based feature extraction. In this paper, we propose an algorithm on estimation. the proposed algorithm, firstly use to estimate 2D face points of target human head, then adopt Relevant Vector Machine evaluate extracted points. Experiments CAS-PEAL-R1 dataset show that has...

Hyperspectral imagery acquired by a hyperspectral sensor contains hundreds of narrow contiguous spectral bands, providing the opportunity to identify various materials present on surface. Due three-dimensional (3D) nature data, 3D filters that could extract joint spatial-spectral features have been recently considered in literature. In this paper, after Gabor with certain orientations extracted from raw image both magnitude and phase used for classification, which is named as Gabor-MP....

Along with the information explosion in Internet era, traditional classification methods, such as KNN (k-nearest neighbor), Naive Bayes (NB), encounter bottlenecks due to endless stream of new words. In this paper, through comparing Rocchio and Bayesian algorithms, it has been found that centroid-based algorithms are insufficient for text classification. Therefore, a novel feature voting model is proposed, which gives rise bag-of-words based algorithm This assigns categories each document...

In this paper, an effective l <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> -norm collaborative representation algorithm based on 3D discrete wavelet transform (3D-DWT) features, called CR_DWT, is proposed for hyperspec-tral image classification. By using the discriminative 3D-DWT features extracted from original spectral space, a non-parametric and efficient CR method developed to calculate coefficients. Due simplicity of method,...

Local binary patterns (LBP) is a simple and efficient texture description operator, the improved "uniform" local (uLBP) has been proved to be powerful feature expression for classification. However, due "0" threshold operation in uLBP, it sensitive gray variance resulting from noises disturbance among classes, which reduces classification performance. In this paper, fuzzy threshold-based uniform patterns, abbreviated as ftuLBP, proposed enhance robustness of descriptor variance....

In recent years, unmanned aerial vehicle technology has been continuously developed and matured, the level of intelligence improved. "Black Flying" brings instability to social security, while UAVs bring convenience routine production daily life. addition, with continuous development UAV cluster technology, future trend air combat will gradually turn cooperation. also be changed from concept reality, theory practice. The outstanding performance artificial in game tasks clearly shows that...

Dimensionality reduction is often adopted for hyperspectral imagery in advance to improve the efficiency of following processing, such as classification and identification. Affinity propagation (AP), which a clustering algorithm, has shown ability automatically pick out representative bands from imagery. Several distance measures have been proposed construct similarity matrix, an important issue AP, but spatial structural information image not considered. In this paper, approach used...