Problem Of Feature Extraction Research Articles

A Non-Contact Fault Diagnosis Method for Rolling Bearings Based on Acoustic Imaging and Convolutional Neural Networks

Rolling bearing fault diagnosis is conventionally performed by vibration-based diagnosis (VBD). However, VBD is restrained in some cases because vibration measurement usually requires the contact with the machine. Acoustical-based fault diagnosis (ABD) has the advantage of non-contact measurement over VBD. However, ABD has received little attention and rarely applied in bearing fault diagnosis. In this paper, a new non-contact ABD method for rolling bearings using acoustic imaging and convolutional neural networks (CNN) is proposed. Firstly, a microphone array is used to acquire the acoustic field radiated by rolling bearings. Then, acoustic imaging is performed with the wave superposition method (WSM). The reconstructed acoustic images can depict the spatial distribution of the acoustic field, which add a new spatial dimension in the acoustic data representation for fault diagnosis and makes it possible to localize the sound sources. Finally, CNN is applied to accomplish bearing fault diagnosis, which can overcome the problems of handcrafted feature extraction in traditional ABD methods. Experimental results verify the effectiveness of the proposed ABD method. Comparisons with peer state-of-the-art ABD methods further validate that the proposed method can mitigate the drawbacks of the existing ABD methods, and obtain more accurate and reliable diagnosis results.

Semi-Supervised Multi-Granularity CNNs for Text Classification: An Application in Human-Car Interaction

With the rapid growth of advanced driving assistance technology, how to implement human-car interaction humanely and personally becomes increasingly important. Existing methods typically rely on hand-crafted data to conduct training, which cannot fully extract information hidden in text. In this paper, we tackle the problem of label data acquisition difficulties and inadequate feature extraction via semi-supervised learning and multi-granularity learning. To fulfill the goal of label data acquisition and feature extraction, we present a novel semi-supervised multi-granularity convolutional neural networks (CNNs)-based (SSMGCNNs-based) model for an application in human-car interaction, which consists of the two-view-embedding (TVE) module and the multi-granularity CNNs (MGCNNs) module. The TVE module learns embeddings of text regions from the unlabeled user command data set and then integrate the learned tv-embeddings into MGCNNs, so that the learned tv-embedded regions are used as an additional input into MGCNNs' convolution layer to solve the problem of data annotation. MGCNNs can fully extract information hidden in text by multiple convolution kernels of the same convolution layer. We compared our model with some state-of-the-art machine learning models. On the car operation command data set, the simulation results demonstrated that, compared with CNNs, our method respectively improved 5.13%, 5.64%, 3.60% and 5.34% on precision, recall, F-1 and training loss.

Semi-GSGCN: Social Robot Detection Research with Graph Neural Network

Malicious social robots are the disseminators of malicious information on social networks, which seriously affect information security and network environments. Efficient and reliable classification of social robots is crucial for detecting information manipulation in social networks. Supervised classification based on manual feature extraction has been widely used in social robot detection. However, these methods not only involve the privacy of users but also ignore hidden feature information, especially the graph feature, and the label utilization rate of semi-supervised algorithms is low. Aiming at the problems of shallow feature extraction and low label utilization rate in existing social network robot detection methods, in this paper a robot detection scheme based on weighted network topology is proposed, which introduces an improved network representation learning algorithm to extract the local structure features of the network, and combined with the graph convolution network (GCN) algorithm based on the graph filter, to obtain the global structure features of the network. An end-to-end semi-supervised combination model (Semi-GSGCN) is established to detect malicious social robots. Experiments on a social network dataset (cresci-rtbust-2019) show that the proposed method has high versatility and effectiveness in detecting social robots. In addition, this method has a stronger insight into robots in social networks than other methods.

Face Recognition based on Linear Projective Non-negative Matrix Factorization with Kullback-Leibler Divergence

In order to solve the problem of feature extraction and classification of Non-negative Matrix Factorization (NMF) in face recognition, a Linear Projective Non-negative Matrix Factorization method based on Kullback-Leibler divergence (LP-NMF-DIV) is proposed. In LP-NMF-DIV, an objective function of Kullback-Leibler divergence is considered. The Taylor series expansion and the Newton iteration formula are used to find the root formula. The iterative algorithm of basis matrix and linear transformation matrix are derived, and the convergence of the algorithm is proved. The results of experimental show that the algorithm is convergent. In face recognition, compared with some typical NMF methods, the algorithm has high recognition accuracy; when the ranks of the basis matrices are set to different values, the algorithm is stable. The method for LP-NMF-DIV is effective.

Low rank theory-based interframe forgery detection for blurry video

It is difficult to extract appropriate features of blurry digital video because of the low quality image for which traditional forgery detection methods are not effective. The video is often blurred due to various complicated conditions, such as limited bandwidth or low pixel complementary metal oxide semiconductor sensors. To address the feature extraction problem for blurry video, we explore one low rank theory to deblur video, which fuses multiple fuzzy kernels of key frame images by low rank decomposition. Further, we extract mean structural similarity features and information fidelity criterion features sequentially by the double detection mechanism to detect forgery points on blurred video frames. We can determine the forgery form of the video and locate the source replication area of the copy-move forgery. Experimental evaluation on two public video databases and videos that we recorded show that this method is robust in blurry video forgery detection, and compared with traditional video forgery detection methods, the efficiency is improved.

Feature Engineering of Solid‐State Crystalline Lattices for Machine Learning

AbstractThe problem of feature extraction, in crystalline solid‐state systems with point defects, is considered. Novel methods for creating features for use in machine‐learning‐based predictive modeling of such systems are developed. The methods are illustrated in a case study where machine learning methods are used to predict the onset of amorphization in crystalline systems containing vacancy defects. How the methods developed may be generalized to study other problems in solid‐state materials is also discussed.

PolSAR Feature Extraction Via Tensor Embedding Framework for Land Cover Classification

Polarimetric synthetic aperture radar (PolSAR) as a typical multi-channel sensor can obtain refined geometrical and geophysical information. In the PolSAR land cover classification task, feature extraction is regarded as a critical step for the final classification. It can employ multi-modal features from the original scattering data, polarimetric target decomposition, and other transformation space. Then, how to efficiently combine multi-modal polarimetric information and extract discriminant features is an important challenge for PolSAR image processing. Graph embedding methods have become a significant technique to deal with feature extraction and dimensionality reduction (DR) problems in recent years. It provides a unified linearization framework in machine learning and other pattern recognition tasks. In this article, an extended tensor embedding framework is introduced to extract the intrinsic features for PolSAR land cover classification. First, each pixel is represented by a feature cube that is constructed by groups of polarimetric scattering signals and target decomposition features in a fixed size patch. Second, an intrinsic matrix is constructed to describe the original geometrical and statistical properties of the samples, and a penalty matrix is designed to represent some constraints. Third, the vector-based algorithms are transformed into tensor space in an unified framework and based on the pair of matrices to obtain the projection matrices in each mode by an iterative optimization process. The effectiveness of the proposed methods is demonstrated on three RADARSAT2 data sets covering the regions of Xi’an, San Francisco, and Flevoland, respectively. The visualization and quantification results show that the proposed method has superiority in land cover classification.

Hybrid Tile Based Feature Extraction and Support Vector Machine Base Content-Based Image Retrieval System for Medical Application

Through the landing of therapeutic endoscopes, earth perception satellites and individual telephones, content-based picture recovery (CBIR) has concerned critical consideration, activated by its broad applications, e.g., medicinal picture investigation, removed detecting, and individual re-distinguishing proof. Be that as it may, developing successful component extraction is as yet reported as an invigorating issue.In this paper, to overcome the feature extraction problems a hybrid Tile Based Feature Extraction (TBFE) is introduced. The TBFE algorithm is hybrid with the local binary pattern (LBP) and Local derivative pattern (LDP). These hybrid TBFE feature extraction method helps to extract the color image features in automatic manner. Support vector machine (SVM) is used as a classifier in this image retrieval approach to retrieve the images from the database. The hybrid TBFE along with the SVM classifier image retrieval is named as IR-TBFE-SVM. Experiments show that IR-TBFE-SVMdelivers a higher correctness and recall rate than single feature employed retrieval systems, and ownsdecentweight balancing and query efficiency performance.

Magnetic Anomaly Target Recognition Based on SVD and SVMs

For the problem of signal feature extraction detected by the magnetic anomaly detection equipment based on the flux gate array, we propose a method of extracting magnetic anomaly signal features by singular value decomposition (SVD) and classifying the magnetic targets using support vector machines (SVMs). The detected magnetic anomaly signal is the singular value decomposed as a 2-D real matrix. An orthogonal basis of a 2-D complete feature space is constructed by the direct product of its left eigenvector matrix and right eigenvector matrix. The coefficient of intersection is the eigenvalue of the magnetic anomaly matrix and can be used as a power density of the magnetic anomaly in its feature space. At the same time, magnetic anomaly signals are reduced continuously to the pole for obtaining high-quality signals and a more complete sample library. Subsequently, nine classification features, such as magnetic gradient tensor elements, were analyzed and selected. The orthogonal basis coefficients of these nine feature matrices were calculated and constructed as feature vectors. The ferromagnetic targets were identified by SVMs when the feature vectors are used. Finally, through experiments, the recognition accuracy rate reaches 90%, and a good target recognition effect is achieved.

Research on multifractal dimension and improved gray relation theory for intelligent satellite signal recognition

In the wake of the development and advancement of signal processing technology for communication radiation source individual, Signal fingerprint feature extraction and analysis technology for communication radiation source individual has broad application prospects in many fields. To effectively extract the individual characteristics of different modulated signals under low SNR environment, and recognize the subtle features for communication radiation source individuals has been a hot spot. Aiming at the problem of signal feature extraction and classifier design under low SNR environment, in the paper, a multifractal dimension and improved gray relation theory based classifier design algorithm is proposed. Firstly, the multifractal dimension feature extraction of nine modulated communication signals is realized. Then multifractal dimension features of these modulated signals under different SNR are compared. An improved gray relation algorithm is used to recognize the extracted subtle characteristics. Meanwhile, FSK signal is used to simulate radio subtle features by adding different distribution of noise. Subtle feature extraction by means of multifractal dimension algorithm and pattern recognition by means of improved gray relation algorithm are used to test the effectiveness of the proposed method for identification of modulated signals and radio subtle features. The simulation results show that the recognition success rate of nine different communication modulated signals can reach 93% even under the SNR of 2 dB, and the recognition success rate of the subtle features of distributed noise can reach 100%. The proposed method provides an effective theoretical basis for identifying of radio modulated signals and communication radiation source individual subtle features.

Effective person re-identification by self-attention model guided feature learning

Person re-identification (re-ID), of which the goal is to recognize person identities of images captured by non-overlapping cameras, is a challenging topic in computer vision. Most existing person re-ID methods conduct directly on detected objects, which ignore the space misalignment caused by detectors, human pose variation, and occlusion problems. To tackle the above mentioned difficulties, we propose a self-attention model guided deep convolutional neural network(DCNN) to learn robust features from image shots. Kernels of the self-attention model evaluate weights for the importance of different person regions. To solve the local feature dependence problem of feature extraction, the non-local feature map generated by the self-attention model is fused with the original feature map generated from the resnet-50. Furthermore, the loss function considers both the cross-entropy loss and the triplet loss in the training process, which enables the network to capture common characteristics within the same individuals and significant differences between distinct persons. Extensive experiments and comparative evaluations show that our proposed strategy outperforms most of the state-of-the-art methods on standard datasets: Market-1501, DukeMTMC-reID, and CUHK03.

Improvement of Segmentation Performance for Feature Extraction on Whirlwind Cloud-based Satellite Image using DBSCAN Clustering Algorithm

Images contain a lot of information that can be used in a variety of areas. One of the images that have much information inside is satellite image. In order to extract the information properly, the image processing step should be performed properly. The segmentation process plays an important role in image processing, especially for feature extraction. Many ways were developed to perform the segmentation image. In this study, we apply DBSCAN clustering to segment images on whirlwind cloud feature extraction problems. DBSCAN is a density-based classifier method which means it is suitable to group a density-based data. While the image used in the segmentation process is the Himawari 8 satellite image which also contains density-based data. It contains various information about clouds condition like cloud type, cloud temperature, cloud humidity, rainfall potential based on cloud temperature, etc. This study uses Himawari 8 satellite images as input where the images taken are images several hours before a wirlwind event in an area, while the cluster method used is the DBSCAN algorithm. Clustering is done to get the extraction features of a wirlwind in the form of centroid points that characterize the movement of a cloud. Segmentation performance was observed based on the number of centroid points as a result of clustering several types of clouds in an area before a wirlwind occurred. Based on segmentation testing using the DBSCAN algorithm for cloud data in an area for several hours before a wirlwind, better segmentation performance was obtained compared to the segmentation results of the Meng hee heng k-means algorithm for the same test data specifications. DBSCAN separates a type of cloud in more detail that makes it easier to record each centroid of each cluster around the scene. It is even able to cluster small groups of clouds independently so that these small groups of clouds can also be detected as features.

A New Feature Extraction Method for Bearing Faults in Impulsive Noise Using Fractional Lower-Order Statistics

According to the performance degradation problem of feature extraction from higher-order statistics in the context of alpha-stable noise, a new feature extraction method is proposed. Firstly, the nonstationary vibration signal of rolling bearings is decomposed into several product functions by LMD to realize signal stability. Then, the distribution properties of product functions in the time domain are discussed by the comparison of heavy tails and characteristic exponent estimation. Fractional lower-order p-function optimization is obtained by the calculation of the distance ratio based on K-means algorithms. Finally, a fault feature dataset is established by the optimal FLOS and lower-dimensional mapping matrix of covariation to accurately and intuitively describe various bearing faults. Since the alpha-stable noise is effectively suppressed and state described precisely, the presented method has shown better performance than the traditional methods in bearing experiments via fractional lower-order feature extraction.

Degradation assessment of rolling bearing towards safety based on random matrix single ring machine learning

Aiming at problems of feature extraction of monitoring data may lose useful information and different feature extraction methods have great influence on the results in rolling bearing performance degradation assessment. A method which is based on random matrix theory to assess the degradation of rolling bearing is proposed to improve production safety. Firstly, the observation matrix was constructed by the raw data of rolling bearing health monitoring, and the random matrix model of rolling bearing operating states was established by the steps of normalization, singular matrix singular value decomposition and standardization. Then, three performance degradation indicators, the average spectral radius, the maximum eigenvalue, and the number of random points in the inner ring are proposed by constructing the linear eigenvalue statistic of the random matrix, combined with the single ring curve of the bearing operating states, to describe the fluctuation discipline of the random matrix elements and the variation law of the characteristic root distribution, thus characterize the statistical random characteristics of the operating states of the rolling bearing as a whole. Finally, the method was applied by the rolling bearing life test vibration data. And four different operation states of the bearing are selected to build the random matrix model and analysis. The analysis results are calculated by single ring curves and quantitative indicators of different states and proved that the method proposed in this paper can effectively identify different degradation states and realize safety states of rolling bearing, meanwhile, it can quantitatively assess the degree of performance degradation.

Cognitive Action Laws: The Case of Visual Features.

This paper proposes a theory for understanding perceptual learning processes within the general framework of laws of nature. Artificial neural networks are regarded as systems whose connections are Lagrangian variables, namely, functions depending on time. They are used to minimize the cognitive action, an appropriate functional index that measures the agent interactions with the environment. The cognitive action contains a potential and a kinetic term that nicely resemble the classic formulation of regularization in machine learning. A special choice of the functional index, which leads to the fourth-order differential equations-Cognitive Action Laws (CAL)-exhibits a structure that mirrors classic formulation of machine learning. In particular, unlike the action of mechanics, the stationarity condition corresponds with the global minimum. Moreover, it is proven that typical asymptotic learning conditions on the weights can coexist with the initialization provided that the system dynamics is driven under a policy referred to as information overloading control. Finally, the theory is experimented for the problem of feature extraction in computer vision.

A Fault Diagnosis Approach for Rolling Bearing Based on Convolutional Neural Network and Nuisance Attribute Projection under Various Speed Conditions

Intelligent fault diagnosis is a promising tool for processing mechanical big data. It can quickly and efficiently process the collected signals and provide accurate diagnosis results. However, rotating machinery often works under various speed conditions, which makes it difficult to extract fault features. Inspired by speech recognition, the nuisance attribute projection method in speech recognition is introduced into fault diagnosis to solve the problem of feature extraction in variable speed signals. Based on the idea of unsupervised feature learning, the loss function of nuisance attribute projection is added to the loss function of convolutional neural network (CNN) to learn fault features from original data. Health status is classified according to the learned characteristics and projection matrix P. A special designed bearing dataset is employed to verify the effectiveness of the proposed method. The results show that the proposed method has a higher accuracy and a simpler framework, which is superior to the existing methods in bearing fault diagnosis.

ECG Signal Classification Algorithm Based on Fusion Features

Cardiovascular disease is a common and frequently-occurring disease in clinic. Cardiovascular disease detection has become the primary concern of people’s health and the mortality is very high. In order to improve the accuracy and universality of the current ECG signal classification algorithm under multi-classification conditions, an ECG signal classification algorithm based on fusion features is proposed. The algorithm uses multi-layer Convolution Neural Network and Recurrent Neural Networks. The spatial and temporal features of ECG signals are extracted automatically. Finally, the accurate classification of ECG signals is realized by using multi-fusion features. In order to verify the accuracy of the algorithm, using MIT-BIH arrhythmia database as the standard test data, the classification accuracy of the algorithm is 99.42%. Compared with the ECG classification algorithm in document[1] [2], the algorithm proposed in this paper abandons the complexity of the traditional ECG algorithm in the process of data feature extraction and solving the problem of insufficient feature extraction of single convolution neural network for one-dimensional ECG signal leads to poor classification effect of small probability samples. The classification sensitivity of ECG classification is improved from 83.00% of single convolution network to 97.07% of class S (supraventricular abnormal beats) classification sensitivity, The classification sensitivity of Class F (Fused Rhythm) is increased from 74.48% to 97.24%, which significantly improves the classification sensitivity of small probability abnormal rhythms, and increases the universality of ECG classification algorithm in multi-class beat recognition.

Hyperspectral image classification with SVM and guided filter

Hyperspectral image (HSI) classification has been long envisioned in the remote sensing community. Many methods have been proposed for HSI classification. Among them, the method of fusing spatial features has been widely used and achieved good performance. Aiming at the problem of spatial feature extraction in spectral-spatial HSI classification, we proposed a guided filter-based method. We attempted two fusion methods for spectral and spatial features. In order to optimize the classification results, we also adopted a guided filter to obtain better results. We apply the support vector machine (SVM) to classify the HSI. Experiments show that our proposed methods can obtain very competitive results than compared methods on all the three popular datasets. More importantly, our methods are fast and easy to implement.

Influence of normalization and color features on super-pixel classification: application to cytological image segmentation

Super-pixel feature extraction is a key problem to get an acceptable performance in color super-pixel classification. Given a color feature extraction problem, it is necessary to know which is the best approach to solve this problem. In the current work, we're interested in the challenge of nucleus and cytoplasm automatic recognition in the cytological image. We propose an automatic process for white blood cells (WBC) segmentation using super-pixel classification. The process is divided into five steps. In first step, the color normalization is calculated. The super-pixels generation by Simple Linear Iterative Clustering algorithm is performed in the second step. In third step, the color property is used to achieve illumination invariance. In fourth step, color features are calculated on each super-pixel. Finally, supervised learning is realized to classify each super-pixel into nucleus and cytoplasm region. The present work rallied an exhaustive statistical evaluation of a very wide variety of the color super-pixel classification, with height normalization methods, four-color spaces and four feature extraction techniques. Normalization and color spaces slightly increase the average accuracy of super-pixel classification. Our experiments based to statistical comparison allow to conclude that comprehensive gray world normalized normalization is better than without normalization for super-pixel classification achieving the first positions in the Friedman ranking. RGB space is the best color spaces to be used in super-pixel feature extraction for nucleus and cytoplasm segmentation. For feature extraction, the learning methods work better on the first order statistics features for the automatic WBC segmentation.

Research on pedestrian tracking algorithm based on deep learning framework

Pedestrian tracking is a hot topic in the field of computer vision. Current pedestrian tracking methods may treat the reappear target as a new target to tracking, which can lead to tracking failure. For this purpose, this paper uses the deep learning framework to complete the whole algorithm of pedestrian tracking to solve this problem. The pedestrian tracking algorithm is divided into pedestrian detection and tracking. In the detection part, the Faster-RCNN framework is used to detect pedestrians; and in the tracking part, we use the Person-ReID method to track pedestrians, converted the pedestrians tracking problem to feature extraction and matching problem between different frames, and improved the tracking results effectively according to the deep learning framework. According to the experiment results on the simple standard dataset and RGB-D People dataset, the tracking mAP of our algorithm get 92.51% and 76.9%.