Components Of Feature Vector Research Articles

Bridge Cable Anomaly Detection Based on Local Variability in Feature Vector of Monitoring Group Cable Forces

The cable is the key supporting component of the cable-stayed bridges, and damaged cables will directly affect the safety and stability of the bridge in operation. Therefore, this paper proposes a bridge cable anomaly detection and localization method based on the variation in the group cable force feature vector. First, mechanical analysis was carried out on the correlation of cable forces between two cables on the same side induced by the single-vehicle case and a cable force feature vector was established to reflect the mechanical characteristics of the group cable forces. Second, a bridge cable anomaly detection method was proposed based on the variation in the group cable force feature vector. Third, the isolation of abnormal components of the cable force feature vector, which can accurately localize abnormal cables, was presented. The long-term monitoring data of an in-service cable-stayed bridge were utilized to validate the effectiveness of the proposed method. The results demonstrated that the proposed method could effectively detect abnormal cable conditions and accurately localize abnormal cables. In addition, the proposed bridge cable feature index and anomaly detection method were robust to the situation of partially missing data due to sensor fault.

A high frequency stock price prediction model based on Hermite basis expansion and LSTM neural network

Based on Hermite basis function expansion and ensemble learning, an improved LSTM neural network method is proposed in this paper. The proposed method can be used for high frequency stock price prediction. Considering the characteristics of high frequency time series data such as high dimensionality, large noise and instability, etc. In this paper, the function information extracted from the Hermite basis function expansion is used to predict the residual sequence predicted by the LSTM neural network. Since the components of the function feature vector are unknown to the underlying model structure of the response variable, the proposed method is processed by Bagging framework. It not only captures the structure of the latent model, but also balances the variance and bias of the model. In addition, the number of prediction periods of the LSTM neural network is a hyperparameter, and the model averaging method based on distance covariance weighting is considered in this paper for optimization. The results of actual data analysis show that the proposed method can effectively optimize the prediction accuracy of the LSTM neural network, and has certain robustness. Finally, on the one hand, this optimization framework can be used to improve other time series prediction models. On the other hand, the proposed method can play an important role in forecasting problems such as daily average temperature prediction and real-time monitoring of atmospheric environmental quality.

Research on High-frequency stock price prediction based on Chebyshev-Stacking and Weighted LSTM neural network [1

To solve the problem of high-frequency stock price prediction, this paper proposed a prediction model based on Chebyshev-Stacking and a weighted LSTM neural network. The proposed method extracts the function characteristic information of the high-frequency stock price series through Chebyshev orthogonal polynomial basis expansion. Considering that the potential model structure between each component of the function feature vector and the residual sequence predicted by the LSTM neural network is unknown and there is a certain noise, this paper used the Stacking framework to enhance the data and weighed the bias and variance of the prediction model. In addition, since the number of predictor variable periods of the LSTM neural network is a hyperparameter, the model averaging method based on distance covariance is used for optimization. The results of actual data analysis show that the proposed method is significantly better than the original LSTM neural network in terms of mean square error, absolute error, and relative error. By selecting the different number of training sets, the robustness of the improved model is verified. Finally, the proposed method can also be used in practical applications such as daily average temperature prediction, missile trajectory prediction, and real-time monitoring of atmospheric environment quality.

DeepKnuckle: Deep Learning for Finger Knuckle Print Recognition

Biometric technology has received a lot of attention in recent years. One of the most prevalent biometric traits is the finger-knuckle print (FKP). Because the dorsal region of the finger is not exposed to surfaces, FKP would be a dependable and trustworthy biometric. We provide an FKP framework that uses the VGG-19 deep learning model to extract deep features from FKP images in this paper. The deep features are collected from the VGG-19 model’s fully connected layer 6 (F6) and fully connected layer 7 (F7). After applying multiple preprocessing steps, such as combining features from different layers and performing dimensionality reduction using principal component analysis (PCA), the extracted deep features are put to the test. The proposed system’s performance is assessed using experiments on the Delhi Finger Knuckle Dataset employing a variety of common classifiers. The best identification result was obtained when the Artificial neural network (ANN) classifier was applied to the principal components of the averaged feature vector of F6 and F7 deep features, with 95% of the data variance preserved. The findings also demonstrate the feasibility of employing these deep features in an FKP recognition system.

Multiple Distance-Based Coding: Toward Scalable Feature Matching for Large-Scale Web Image Search

For scalable feature matching in large-scale web image search, the bag-of-visual-words-based (BOW) approaches generally code local features as visual words to construct an inverted index file to match features efficiently. Both the popular feature coding techniques, i.e., K-means-based vector quantization and scalar quantization, directly quantize features to generate visual words. K-means-based vector quantization requires expensive visual codebook training, whereas scalar quantization leads to the miss of many matches due to the low stability of individual components of feature vectors. To address the above issues, we demonstrate that the corresponding sub-vectors of similar features generally have similar distances to multiple reference points in feature subspace and propose a multiple distance-based feature coding scheme for scalable feature matching. Specifically, based on the distances between the sub-vectors and multiple distinct reference points, we transform each feature to a set of feature codes, where one code is treated as a visual word required to construct the inverted index file whereas the others are embedded into the index file to further verify the feature matching based on the visual words. The proposed coding scheme does not need visual codebook training and shows desirable stability and discriminability. Moreover, in the matching verification, a feature-distance estimation method is proposed to estimate the Euclidean distances between features for an accurate matching verification. Extensive experimental results demonstrate the superiority of the proposed approach in comparison to the other approaches using recent feature quantization methods for large-scale web image search.

Research on the application of artificial intelligence method in automobile engine fault diagnosis

The application of artificial intelligence methods in fault diagnosis is becoming more and more extensive, and exploring and researching intelligent fault diagnosis methods for automobile engines is also a hot spot in the field of automotive engineering. Different machine learning methods have their own advantages and disadvantages. By extracting different characteristic parameters and optimizing the combination of multiple algorithms, faster and stable diagnosis results can be achieved, so that faults can be discovered and repaired in time. Aiming at the potential fluctuation and impact characteristics of vibration plus signal caused by different failure states of automobile engines, this paper proposes two engine fault identification methods using vibration acceleration signals as diagnostic parameters. They are Cross Validation -Support Vector Machine (CV-SVM)and Particle Swarm Optimization-Probabilistic Neural Network (PSO-PNN) engine fault identification methods. The advantages and disadvantages of the two methods are compared and analyzed. Obtain the amplitude corresponding to the frequency multiplication of the vibration acceleration signal through the spectrum analysis method, which is used as the main component of the input feature vector, and establish the SVM fault diagnosis model combined with the cross-validation method (CV); In addition, multiple one-dimensional arrays composed of time-domain signals are directly used as input feature vectors, and the particle swarm optimization (PSO) parameter optimization is used to obtain the best Probabilistic Neural Network(PNN) fault diagnosis model. The results show that both the CV-SVM (small sample) method and the PSO-PNN method (large sample) can realize the identification and diagnosis of the established engine fault type. The CV-SVM method has more advantages in diagnostic fault tolerance, but the PSO-PNN method can simplify the process of feature sample preparation, save a lot of manual feature extraction tasks, and is more convenient in practical application.

Active Sonar Target Classification with Power-Normalized Cepstral Coefficients and Convolutional Neural Network

Detection and classification of unidentified underwater targets maneuvering in complex underwater environments are critical for active sonar systems. In previous studies, many detection methods were applied to separate targets from the clutter using signals that exceed a preset threshold determined by the sonar console operator. This is because the high signal-to-noise ratio target has enough feature vector components to separate. However, in a real environment, the signal-to-noise ratio of the received target does not always exceed the threshold. Therefore, a target detection algorithm for various target signal-to-noise ratio environments is required; strong clutter energy can lead to false detection, while weak target signals reduce the probability of detection. It also uses long pulse repetition intervals for long-range detection and high ambient noise, requiring classification processing for each ping without accumulating pings. In this study, a target classification algorithm is proposed that can be applied to signals in real underwater environments above the noise level without a threshold set by the sonar console operator, and the classification performance of the algorithm is verified. The active sonar for long-range target detection has low-resolution data; thus, feature vector extraction algorithms are required. Feature vectors are extracted from the experimental data using Power-Normalized Cepstral Coefficients for target classification. Feature vectors are also extracted with Mel-Frequency Cepstral Coefficients and compared with the proposed algorithm. A convolutional neural network was employed as the classifier. In addition, the proposed algorithm is to be compared with the result of target classification using a spectrogram and convolutional neural network. Experimental data were obtained using a hull-mounted active sonar system operating on a Korean naval ship in the East Sea of South Korea and a real maneuvering underwater target. From the experimental data with 29 pings, we extracted 361 target and 3351 clutter data. It is difficult to collect real underwater target data from the real sea environment. Therefore, the number of target data was increased using the data augmentation technique. Eighty percent of the data was used for training and the rest was used for testing. Accuracy value curves and classification rate tables are presented for performance analysis and discussion. Results showed that the proposed algorithm has a higher classification rate than Mel-Frequency Cepstral Coefficients without affecting the target classification by the signal level. Additionally, the obtained results showed that target classification is possible within one ping data without any ping accumulation.

A New Framework Combining Local-Region Division and Feature Selection for Micro-Expressions Recognition

Micro-expressions are deliberate or unconscious movements of people's psychological activities, reflecting the transient facial true expressions. Previous works focus on the whole face for micro-expressions recognition. These methods can extract a number of feature vectors which are relevant or irrelevant to the micro-expressions recognition. Besides, the high-dimension feature vectors can result in longer computational time and increased computational complexity. In order to address these problems, we propose a new framework which combines the local-region division and the feature selection. Based on the proposed framework, the original images can retain more efficient regions and filter out the invalid components of feature vectors. Specifically, with the joint efforts of the facial deformation identification model and facial action coding system, the global region is divided into seven local regions with their corresponding actions units. The ReliefF algorithm is used to select effective components of feature vectors and reduce the dimension. To evaluate the proposed framework, we conduct experiments on both the Chinese Academy of Sciences Micro-expression II Database and Spontaneous Micro-expression Database with Leave-One-Subject-Out Cross Validation method. The results show that the performance in local combined regions outperforms its counterpart in the global region, and the recognition accuracy is further improved with the combination of feature selection.

Underwater Moving Target Classification Using Multilayer Processing of Active Sonar System

The task of detecting and classifying highly maneuverable and unidentified underwater targets in complex environments is significant in active sonar systems. Previous studies have applied many detection schemes to this task using signals above a preset threshold to separate targets from clutter; this is because a high signal-to-noise ratio (SNR) target has sufficient feature vector components to be separated out. However, in real environments, the received target return’s SNR is not always above the threshold. Therefore, a target detection algorithm is needed for varied target SNR conditions. When the clutter energy is too strong, false detection can occur, and the probability of detection is reduced due to the weak target signature. Furthermore, since a long pulse repetition interval is used for long-range detection and ambient noise tends to be high, classification processing for each ping is needed. This paper proposes a multilayer classification algorithm applicable to all signals in real underwater environments above the noise level without thresholding and verifies the algorithm’s classification performance. We obtained a variety of experimental data by using a real underwater target and a hull-mounted active sonar system operated on Korean naval ships in the East Sea, Korea. The detection performance of the proposed algorithm was evaluated in terms of the classification rate and false alarm rate as a function of the SNR. Since experimental environment data, including the sea state, target maneuvering patterns, and sound speed, were available, we selected 1123 instances of ping data from the target over all experiments and randomly selected 1000 clutters based on the distribution of clutters for each ping. A support vector machine was employed as the classifier, and 80% of the data were selected for training, leaving the remaining data for testing. This process was carried out 1000 times. For the performance analysis and discussions, samples of scatter diagrams and feature characteristics are shown and classification tables and receiver operation characteristic (ROC) curves are presented. The results show that the proposed algorithm is effective under a variety of target strengths and ambient noise levels.

Feature vector clustering molecular pairs in computer simulations.

A clustering framework is introduced to analyze the microscopic structural organization of molecular pairs in liquids and solutions. A molecular pair is represented by a representative vector (RV). To obtain RV, intermolecular atom distances in the pair are extracted from simulation trajectory as components of the key feature vector (KFV). A specific scheme is then suggested to transform KFV to RV by removing the influence of permutational molecular symmetry on the KFV as the predicted clusters should be independent of possible permutations of identical atoms in the pair. After RVs of pairs are obtained, a clustering analysis technique is finally used to classify all the RVs of molecular pairs into the clusters. The framework is applied to analyze trajectory from molecular dynamics simulations of an ionic liquid (trihexyltetradecylphosphonium bis(oxalato)borate ([P6,6,6,14 ][BOB])). The molecular pairs are successfully categorized into physically meaningful clusters, and their effectiveness is evaluated by computing the product moment correlation coefficient (PMCC). (Willett, Winterman, and Bawden, J. Chem. Inf. Comput. Sci. 1986, 26, 109-118; Downs, Willett, and Fisanick, J. Chem. Inf. Comput. Sci. 1994, 34, 1094-1102) It is observed that representative configurations of two clusters are related to two energy local minimum structures optimized by density functional theory (DFT) calculation, respectively. Several widely used clustering analysis techniques of both nonhierarchical (k-means) and hierarchical clustering algorithms are also evaluated and compared with each other. The proposed KFV technique efficiently reveals local molecular pair structures in the simulated complex liquid. It is a method, which is highly useful for liquids and solutions in particular with strong intermolecular interactions. © 2019 Wiley Periodicals, Inc.

Identification of location and size of a defect in a structural system employing active external excitation and hybrid feature vector components in HMM

For the fault diagnosis of a mechanical system, various kinds of methods have been developed so far. For a structural system having a defect, pattern recognition methods such as Hidden Markov model (HMM) and Artificial neural network (ANN) are widely used in engineering fields. A statistical model can be constructed with one of the methods using various signals that are extracted from the structural system of interest. In the present study, a HMM employing hybrid feature vector measures is proposed for the fault diagnosis of a structural system having a defect. To obtain the hybrid feature vector components, five frequency response peaks obtained with FFT and two additional components obtained with ANN are employed. For the proposed method, an active external excitation having some specific frequency components is also applied to the structure to overcome the noise effect. To verify the effectiveness of the proposed method, a numerical model of a rotating blade having a crack is employed. Acceleration signals extracted from the structural system are employed to develop the proposed model so that the location and size of the crack can be identified. Using the proposed method, the diagnostic accuracy of the identification is significantly improved even with high level of noise in the system.

A Family of Shape Ellipticity Measures for Galaxy Classification

A new family of shape ellipticity measures is introduced. Each measure from the family distinguishes among the ellipses with different axis length ratios. This is not case for existing ellipticity measures. The new measures are theoretically well founded, which helps us to better understand their behavior and suitability to certain applications. All measures from the family are invariant with respect to translation, rotation, and scaling transformations; range over $(0,1]$; and pick the value 1 only for the ellipses with a specific ratio between minor and major axis lengths. The measures from the new family are applied to the galaxy classification tasks. The 100% classification rate was achieved by using a $k$-NN classifier. Only six parameters (feature vector components), computed from three different ellipticity measures, were used. Based on 100 mutually independent experiments, an average classification rate of 95.6% was obtained. Previously accuracies of 92.3% and 95.1% were obtained using nearest neighbor and neural network classifiers, respectively [S. Lekshmi, K. Revathy, and S. R. Prabhakaran Nayar, Astron. Astrophys., 405 (2003), pp. 1163--1167].

Emotion recognition improvement using normalized formant supplementary features by hybrid of DTW-MLP-GMM model

In recent four decades, enormous efforts have been focused on developing automatic speech recognition systems to extract linguistic information, but much research is needed to decode the paralinguistic information such as speaking styles and emotion. The effect of using first three normalized formant frequencies and pitch frequency as supplementary features on improving the performance of an emotion recognition system that uses Mel-frequency cepstral coefficients and energy-related features, as the components of feature vector, is investigated in this paper. The normalization is performed using a dynamic time warping-multi-layer perceptron hybrid model after determining the frequency range that is most affected by emotion. To reduce the number of features, fast correlation-based filter and analysis of variations (ANOVA) methods are used in this study. Recognizing of the emotional states is performed using Gaussian mixture model. Experimental results show that first formant (F1)-based warping and ANOVA-based feature selection result in the best performance as compared to other simulated systems in this study, and the average emotion recognition accuracy is acceptable as compared to most of the recent researches in this field.

Down Syndrome Diagnosis Based on Gabor Wavelet Transform

Down syndrome is a chromosomal condition caused by the presence of all or part of an extra 21st chromosome. It has different facial symptoms. These symptoms contain distinctive information for face recognition. In this study, a novel method is developed to distinguish Down Syndrome in a custom face database. Gabor Wavelet Transform (GWT) is used as a feature extraction method. Dimension reduction is performed with Principal Component Analysis (PCA). New dimension which has most valuable information is derived with Linear Discriminant Analysis (LDA). Classification process is implemented with k-nearest neighbor (kNN) and Support Vector Machine (SVM) methods. The classification accuracy is carried out 96% and 97,34% with kNN and SVM methods, respectively. Different from the studies related with the Down Sydrome, feature selection process is applied before PCA according to the correlation between components of feature vectors. Best results are achieved with euclidean distance metric for kNN and linear kernel type for SVM. In this way, we developed an efficient system to recognize Down syndrome.

Multi-stream LSTM-HMM decoding and histogram equalization for noise robust keyword spotting

Highly spontaneous, conversational, and potentially emotional and noisy speech is known to be a challenge for today's automatic speech recognition (ASR) systems, which highlights the need for advanced algorithms that improve speech features and models. Histogram Equalization is an efficient method to reduce the mismatch between clean and noisy conditions by normalizing all moments of the probability distribution of the feature vector components. In this article, we propose to combine histogram equalization and multi-condition training for robust keyword detection in noisy speech. To better cope with conversational speaking styles, we show how contextual information can be effectively exploited in a multi-stream ASR framework that dynamically models context-sensitive phoneme estimates generated by a long short-term memory neural network. The proposed techniques are evaluated on the SEMAINE database-a corpus containing emotionally colored conversations with a cognitive system for "Sensitive Artificial Listening".

Dimension reduction by a novel unified scheme using divergence analysis and genetic search

In this study, a unified scheme using divergence analysis and genetic search is proposed to determine significant components of feature vectors in high-dimensional spaces, without having to deal with singular matrix problems. In the literature it is observed that three main problems exist in the feature selection process performed in a high-dimensional space. These problems are high computational load, local minima, and singular matrices. In this study, feature selection is realized by increasing the dimension one by one, rather than reducing the dimension. In this sense, the recursive covariance matrices are formulated to decrease the computational load. The use of genetic algorithms is proposed to avoid local optima and singular matrix problems in high-dimensional feature spaces. Candidate strings in the genetic pool represent the new features formed by increasing the dimension. The genetic algorithms investigate the combination of features which give the highest divergence value. In this study, two methods are proposed for the selection of features. In the first method, features in a high-dimensional space are determined by using divergence analysis and genetic search (DAGS) together. If the dimension is not high, the second method is offered which uses only recursive divergence analysis (RDA) without any genetic search. In Section 3 two experiments are presented: Feature determination in a two-dimensional phantom feature space, and feature determination for ECG beat classification in a real data space.

Modeling the functional consequences of single residue replacements in bacteriophage f1 gene V protein

A computational mutagenesis methodology utilizing a four-body, knowledge-based, statistical contact potential is applied toward globally quantifying relative environmental perturbations (residual scores) in bacteriophage f1 gene V protein (GVP) due to single amino acid substitutions. We show that residual scores correlate well with experimentally measured relative changes in protein function upon mutation. Residual scores also distinguish between GVP amino acid positions grouped according to protein structural or functional roles or based on similarities in physicochemical characteristics. For each mutant, the in silico mutagenesis additionally yields local measures of environmental change (EC scores) occurring at every residue position (residual profile) relative to the native protein. Implementation of the random forest (RF) algorithm, utilizing experimental GVP mutants whose feature vector components include EC scores at the mutated position and at six structurally nearest neighbors, correctly classifies mutants based on function with up to 77% cross-validation accuracy while achieving 0.82 area under the receiver operating characteristic curve. A control experiment highlights the effectiveness of mutant feature vector signals, and a variety of learning curves are generated to analyze the impact of GVP mutant data set size on performance measures. An optimally trained RF model is subsequently used for inferring function for all the remaining unexplored GVP mutants.

ＤＣＴによる次元圧縮と事例選択を用いたビデオ超解像アルゴリズムの高速化

In learning-based super-resolution algorithms, there are two major problems. One is that they require a large amount of memory to store examples; the other is the high computational cost of finding the nearest neighbors in the database. We have developed a novel learning-based video super-resolution algorithm with less memory requirements and computational cost. To this end, we adopted discrete cosine transform coefficients for feature vector components. Moreover, we designed an example selection procedure to construct a compact database. We conducted evaluative experiments using MPEG test sequences and real images to synthesize a high-resolution video. Experimental results show that our method improves the effectiveness of super-resolution algorithms, while preserving the quality of synthesized images.

An FPGA-based accelerator for Fourier Descriptors computing for color object recognition using SVM

Fourier Descriptors (FD) can be used as feature vector components in various applications, such as real-time color object recognition or image retrieval. The full process is composed of the feature extraction followed by a classification step performed using support vector machine (SVM). In order to accelerate the computation of FD, a hardware implementation using FPGA technology is presented in this paper. We evaluated classification performance with respect to lighting variations and noise sensibility. Several experiments were carried out on three databases. Then an efficient architecture for FD computation on FPGAs is proposed and designed as accelerator. The WildCard is used to prototype this implementation. This design can have an operation speed up of approximately 10 compared to the standard software PC implementation.

Method of noise reduction based on dynamic aspects of speech

A system and method are provided that reduce noise in pattern recognition signals. To do this, embodiments of the present invention utilize a prior model of dynamic aspects of clean speech together with one or both of a prior model of static aspects of clean speech, and an acoustic model that indicates the relationship between clean speech, noisy speech and noise. In one embodiment, components of a noise-reduced feature vector are produced by forming a weighted sum of predicted values from the prior model of dynamic aspects of clean speech, the prior model of static aspects of clean speech and the acoustic-environmental model.