362 results about "Network classification" patented technology

An artificial neural network device that utilizes one or more non-volatile memory arrays as the synapses. The synapses are configured to receive inputs and to generate therefrom outputs. Neurons are configured to receive the outputs. The synapses include a plurality of memory cells, wherein each of the memory cells includes spaced apart source and drain regions formed in a semiconductor substrate with a channel region extending there between, a floating gate disposed over and insulated from a first portion of the channel region and a non-floating gate disposed over and insulated from a second portion of the channel region. Each of the plurality of memory cells is configured to store a weight value corresponding to a number of electrons on the floating gate. The plurality of memory cells are configured to multiply the inputs by the stored weight values to generate the outputs.

The invention discloses a system and a method for recognizing a remote sensing image target based on migration network learning, mainly solving the problems that the correct recognition rate for a remote sensing image with a label is relatively low when the number of data is less and the obtaining of the image label is difficult and needs high cost in the conventional methods. The whole system comprises an image characteristic extracting module, a migration network classifier learning system generating module and a migration network classifier learning system learning module, wherein the image characteristic extracting module is used for completing the characteristic extraction of the image; the migration network classifier learning system generating module is used for training input sample data by a network integrated learning algorithm introduced into migration learning to obtain a migration network classifier learning system; and the migration network classifier learning system learning module is used for completing the classification and the recognition of the characteristics of a new sample image. The invention has the advantage of the capability of utilizing other existing resources to improve the correct recognition rate of the remote sensing image target without collecting data again and can be used for the target recognition of the remote sensing image.

The invention provides a large-amplitude face straightening method by means of an adversarial network and a three-dimensional morphological model. The main content of the large-amplitude face straightening method by means of an adversarial network and a three-dimensional morphological model includes a reconstruction module, a generation network and classification module, and an identification module. The large-amplitude face straightening method by means of an adversarial network and a three-dimensional morphological model includes the steps: a generator generates a forward front face image by taking a non-forward front face image as input, and at the same time a classifier tries to determine whether the image is a real image and utilizes the fed back information to promote the image generated from the generator to be more close to the real image, and at the same time an identification engine is used to maintain the original identity characteristics in the input image. The large-amplitude face straightening method by means of an adversarial network and a three-dimensional morphological model can process the non forward face, especially a large-amplitude deflected face image, can provide a generation network and a morphological model to straighten the face, and can greatly improve the effect of face identification and straightening at the same time.

Partial pre-encryption with network-based packet sorting. A video-on-demand (VOD) delivery system for delivering encrypted transport streams to incumbent and overlay set-top boxes utilizes a packet picker / duplicator for sorting selected packets from non-selected packets, duplicating the selected packets, and encrypting one of the pair of duplicated selected packets according to an incumbent encryption scheme. A VOD file server stores the transport stream from the packet picker / duplicator. A network sorter sorts the unencrypted selected packet from the non-selected packets and the encrypted selected packet. The network sorter also sorts the encrypted selected packet from the non-selected packets and encrypts the unencrypted selected packets and the non-selected packets according to an overlay encryption scheme and then sends the transport stream to an overlay set-top box. The network sorter is also responsible for combining the non-selected packets and the incumbent encrypted packets and to send the transport stream to an incumbent set-top box.

The invention discloses a deep convolutional neutral network and superpixel-based image semantic segmentation method, which overcomes the problem that the precision of an existing semantic segmentation method still needs to be improved in the prior art. The method comprises the following steps of 1, training a deep convolutional neutral network classification model from images to category labels on an image classification data set; 2, adding a deconvolutional layer to the deep convolutional neutral network classification model, performing fine adjustment training on an image semantic segmentation data set, and realizing mapping from images to image semantic segmentation results; 3, inputting test images to a deep convolutional neutral network semantic segmentation model to obtain semantic labels of pixels, and inputting the test images to a superpixel segmentation algorithm to obtain a plurality of superpixel regions; and 4, fusing superpixels and the semantic labels to obtain a final improved semantic segmentation result. The method improves the precision of the existing semantic segmentation method and is of important significance in image identification and application.

The invention discloses a health condition monitoring method of complex equipment based on an attention mechanism and a neural network. The main steps comprise acquiring tge multi-sensor data of complex equipment; obtaining the effective measurement data by feature selection; obtaining a plurality of slice samples by preprocessing; establishing the neural network classification model which combines the attention mechanism and the depth neural network; inputting the slice samples and their corresponding labels into the neural network classification model to train the neural network classification model offline; inputting the slice samples of multi-sensor data to be predicted into the trained neural network classification model to obtain the health status of complex equipment. The method considers the data characteristics of the multi-sensor signal, fully excavates the local characteristics and the time sequence information in the data, has high prediction accuracy and wide applicability, and can be widely applied to various complex equipment.

Disclosed herein is an apparatus and method for detecting a person from an input video image with high reliability by using gradient-based feature vectors and a neural network. The human detection apparatus includes an image preprocessing unit for modeling a background image from an input image. A moving object area setting unit sets a moving object area in which motion is present by obtaining a difference between the input image and the background image. A human region detection unit extracts gradient-based feature vectors for a whole body and an upper body from the moving object area, and detects a human region in which a person is present by using the gradient-based feature vectors for the whole body and the upper body as input of a neural network classifier. A decision unit decides whether an object in the detected human region is a person or a non-person.

The present application relates to an intelligent outpatient guiding method and device, a computer device and a storage medium. The method includes: acquiring patient inquiry data; extracting inquiryelement features from the patient inquiry data; converting the inquiry element features into feature vectors; inputting the feature vectors into a preset neural network classifier to obtain a first department matching probability; inputting the feature vectors into a preset Bayesian classifier to obtain a second department matching probability; and generating a department recommendation result isaccording to the first department matching probability and the second department matching probability. The method can improve the accuracy of patient registration.

The invention relates to a method for automatic identification and detection of defects in a composite material. The method comprises steps of: detecting the composite material to generate an infrared image by using infrared thermal wave nondestructive testing equipment; conducting phase space reconstruction on the infrared sequence image to determine defect position of the composite material and segment defect area of the image; conducting phase space reconstruction on the infrared sequence image with defect area and carrying out singular value decomposition to obtain a singular matrix, and left and right projection matrixes; carrying out matrix reconstruction again on the two projection matrixes; extracting algebraic characteristics of time information and space information of the defect through singular value decomposition; constructing mixing characteristic vector as characteristic symptom of the defect; and utilizing results from a nerve network classifier to complete the identification and classification determination. The method of the invention can realize automatic identification and detection on defect in the composite material, carry out rapid detection on damage type of the composite material and provide rapid detection means according to usage condition of the composite material, and has critical reality meaning and research value.

Genetically adaptive neural network systems and methods provide environmentally adaptable classification algorithms for use, among other things, in multi-static active sonar classification. Classification training occurs in-situ with data acquired at the onset of data collection to improve the classification of sonar energy detections in difficult littoral environments. Accordingly, in-situ training sets are developed while the training process is supervised and refined. Candidate weights vectors evolve through genetic-based search procedures, and the fitness of candidate weight vectors is evaluated. Feature vectors of interest may be classified using multiple neural networks and statistical averaging techniques to provide accurate and reliable signal classification.

The invention discloses a real-time scene image semantic segmentation method based on a lightweight network, comprising the following steps: S1, training a lightweight network classification model according to a scene image data set; S2, constructing a deep convolution neural network model based on a lightweight network classification model; S3, inputting the training data of the scene image dataset into a depth convolution neural network, output a predicted image, comparing the predicted image with a semantic label image of the scene image data set, and calculating a cross entropy loss as anobjective function to obtain a trained image semantic segmentation model; S4, inputting the real-time scene image to be tested into the image semantic segmentation model to obtain the image semanticsegmentation result. By taking the modified MobileNetV2 as a basic network, the invention can efficiently extract image features, and in the upsampling process, a quick connection block is used, so that the parameter utilization is more efficient, and the speed of the semantic segmentation model is further improved.

The invention provides a cloud data center network deployment scheme based on a virtual router. By introducing the concept of a virtual exchanger and the virtual router which are independently developed, the network concept from the physics aspect is abstracted to the virtual aspect so that unified management can be carried out by webmasters conveniently and the network expense can be saved. The scheme includes the steps that firstly, a basic network environment is built; secondly, the virtual exchanger and the virtual router are deployed; thirdly, subnetting is carried out; fourthly, network services are configured, a network is scientifically partitioned into an external network, an internal network and a route network by in-depth investigation and research on network requirements of customers, and different levels of deployment schemes are made according to different network classifications. Flexibility requirements of customers are better met, at the same time, classified and hierarchical network management is achieved, deployment is more scientific and concise and management is facilitated.

The invention discloses a semantic segmentation method based on double-flow feature fusion. The method comprises the steps of in a training stage, constructing a convolutional neural network comprising an input layer, a hidden layer and an output layer, and the hidden layer comprises an RGB image processing module, a depth image processing module, a fusion module and a first deconvolution layer; inputting the original image into a convolutional neural network for training to obtain a corresponding semantic segmentation prediction graph; calculating a loss function value between a set formed bysemantic segmentation prediction images corresponding to the original images and a set formed by one-hot coded images processed by corresponding real semantic segmentation images to obtain an optimalweight vector and a bias term of a convolutional neural network classification training model; and in a test stage, inputting an indoor scene image to be semantically segmented into the convolutionalneural network classification training model to obtain a predicted semantic segmentation image. According to the method, the semantic segmentation efficiency and accuracy of the indoor scene image are improved.

The invention relates to an urban road traffic condition detection method based on the voting of a network sorter, which comprises the following steps: 1) monitoring traffic characteristic parameters in real time, and extracting the traffic characteristic parameters so as to obtain test sample sets, wherein the traffic characteristic parameters comprises average vehicle speed v(m / s), vehicle flow f(veh / s), time occupancy ratio s, and travel time t(s); 2) respectively carrying out classification by a SVM (support vector machine)classifier, a BP(beeper) neural network, and a SVM-BP (support vector machine-beeper) cascade classifier in the method of combining the voting of the SVM classifier, a BP neural network, and a SVM-BP cascade classifier; if the three categories are same, sorting the test samples into the category; or if two categories are same, comparing the sum of two weights of the same category with the weight of different categories so as to determine large weights as the classification results; or if the three categories are different, taking the results recognized by the classifiers with the highest weights as the results after the mergence. The urban road traffic condition detection method provided by the invention can efficiently enhance the accuracy.

A machine-aided smell identification method based on modularized combination neural net characterized by that, the combination neural net categorizer comprises a forward direction recessive single layer perceptron module and a RBF neural net module, wherein each categorizer module includes a plurality of sub-modules, whose arrangement is determined through the two phases of accretion and trimming, the parameters of center, width and weight value can be determined through error reversion algorithm. The invention solves the problem of the high dimension multiple catalog mass sample set in a fast and effective way. The machine-aided device by the invention can be used to distinguish many kinds of smells and to estimate the odor intensity simultaneously.

The invention relates to a feature fusion coefficient learnable image semantic segmentation method. The method mainly comprises the following steps: to begin with, training a deep convolution networkclassification model from image to category label in an image classification data set; converting a full connection layer type in the classification model into a convolutional layer type to obtain a full convolution deep neural network model for category prediction at the pixel level; expanding convolutional layer branch, and setting a coefficient for each branch, feature fusion layers being fusedaccording to coefficient proportion, and the coefficient being set in a learnable state; then, carrying out fine-tuning training in an image semantic segmentation data set, and meanwhile, carrying out coefficient learning to obtain a semantic segmentation model; carrying out fine-tuning training and fusion coefficient learning to obtain 1-20 groups of fusion coefficients; and finally, selecting the branch, the coefficient of which is largest, from each group, carrying out final combination, and carrying out fine-tuning training and coefficient learning again to obtain a final semantic segmentation model. The method enables the feature fusion effect to reach a best state.

The invention discloses an FNN (Fuzzy Neural Network) learning algorithm which specifically comprises the steps of establishment of an FNN, training of FNN parameters, prediction identification on the trained FNN, and judgment of target category through the actual output result. The FNN learning algorithm provided by the invention has the benefits that through training to the established FNN, a network classifier comprising a weight is obtained, and during training, the learning rate is changed along with the change of gradient in the training process, so that accurate identification on mistakenly identified data is realized, and a method that a counting backward technique is adopted to enable the partial derivative solving to a denominator variable to be changed into the partial derivative solving to a common variable is provided, the amount of operation of the system is reduced, and the efficiency of the system is further improved; finally, the trained FNN is utilized to perform identification prediction on the test data, and compared with the traditional FNN, the identification rate is remarkably improved.

The invention provides an intelligent fault diagnosis method based on a rough Bayesian network classifier, which comprises the following steps: using standard fault feature data as a fault diagnosis condition attribute set, using a standard fault mode as a fault diagnosis decision attribute set, and adopting a rough set principle to construct an original fault diagnosis information table T1; adopting the minimum entropy method to carry out discrete processing on various continuous fault diagnosis condition attribute values in the T1, so as to form a discretization fault diagnosis information table T2; using a rough set discernable matrix and a nuclear theory to carry out attribute reduction and optimal feature selection on the T2, so as to form a reduction fault diagnosis information table T3; and using the T3 to establish the Bayesian network classifier, so as to realize efficient and quick intelligent fault diagnosis. The intelligent fault diagnosis method avoids the 'curse of dimensionality' problem existed in a Bayesian network diagnostic method, overcomes weaknesses of rigid reasoning and critical misjudgment in a rough set diagnostic method, and greatly improves the efficiency and accuracy of fault diagnosis.

The present invention discloses a tree species classification method based on multi-source simultaneous high-resolution remote sensing data. High-resolution and hyperspectral data simultaneously acquired by an integrated sensor is utilized; firstly, coronal breadth identification is carried out based on the high-resolution data and an object-oriented method, then tree species classification is carried out based on spatial details and spectral features, which are extracted by the hyperspectral data, by combining a BP neural network classifier, and finally, the accuracy is verified by a confusion matrix. According to the tree species classification method disclosed by the invention, based on an edge detection multi-scale segmentation method, segmentation grades with different scales are established from multiple levels and multiple patterns and segmentation and information extraction are carried out layer by layer, so that the classification accuracy of tree species and forest types of subtropical natural secondary forests are promoted.

The invention discloses a sample categorization method based on a depth neural network. The open set category mining and extending method based on a depth neural network comprises steps of using a sample set comprising defined category samples to train a categorized model to be extended, obtaining categorization threshold value information, sending a sample set comprising undefined category samples into the categorization model to be extended, determining at least part of the undefined category samples according to the categorization threshold value information of the categorization model to be extended, artificially marking the undefined category samples, adding a number of columns of a weight transfer matrix in a categorization layer of the depth nerve network in order to increase a total number of model recognition categories, wherein the added weight columns comprise first information associated with global categorization and second information associated with connection between categories and using the undefined category samples which are artificially marked to increase the models which already finish training and updating. The open set category mining and extending method based on the depth neural network and the device thereof extend the depth neural network through modifying a depth neural network categorization layer weight transfer matrix, dynamically increases the number of the recognized categories so as to process the open set recognition problem and can be applied to a scene which is closer to a real scene.

The invention discloses a handwritten Uyghur character recognition method based on classifier integration, which belongs to the pattern recognition field and includes steps of: pre-processing handwritten Uyghur characters, extracting feature vectors of the Uyghur characters by aid of directional line elements, respectively using a modified quadratic discriminant function (MODF) classifier and a back propagation (BP) neural network classifier to classify the feature vectors of the Uyghur characters, and integrating classification recognition results, namely an MODF confidence value set and a BP neural network confidence value set, wherein the step of integrating especially includes achieving integrating through the weighted sum of two confidence values to obtain a final confidence value set and selecting the maximum confidence value to serve as the recognition result. Two classifiers are used in the method, and recognition rate is improved due to result integrating. The handwritten Uyghur character recognition method based on classifier integration has the advantages of being good in algorithm performance, strong in instantaneity, high in reliability and high in recognition rate, is mainly applied to a mobile platform, namely a cell phone, to achieve handwritten Uyghur character recognition, lays a foundation for Uyghur character informatization processing, and provides a new method and an application way for Uyghur character cultural exchange.

The present invention relates to an interconnection network classification advertisement electronic business system whose number of object-receiving people can be dependent on advertising fee. It is characterized by that before said advertisement is released, the required number of object-receiving people and classification can be defined according to its actual requirement, the advertisement releaser can pay related fee to system service company in advance, so that the advertisement releaser can use the described system to release advertisement and conveniently establish on-network shop, and can real-timely monitor the conditions of that the advertisement is browsed and its product and service are ordered.

The invention discloses an intelligent graphic retrieval method, which is characterized in that: extracting features of graphics to generate feature set by the method of Fourier change, training RBF neural network classification model by taking one part of the feature set as training set, indexing the graphics using the classification result given by the classification model; client of a retrieval system extracts the features of retrieval graphics, gives a category by the trained classification model and computes the similarity distance between the retrieval graphics and each graphics in the feature set of the same category; sorting the similarity distance, returning to the graphics according to the number made by the system and further revising RBF neural network classification model by relevant feedback methods. The invention improves the intelligence of search process, effectively determines the RBF neural network classification model by improved algorithm of subtractive clustering, greatly improves retrieval precision, speeds up retrieval speed and upgrades retrieval performance.

The invention provides a method for diagnosing faults of urban gas pipelines based on a deep learning neural network. According to the method, a deep learning neural network classification model formed by a sparse automatic encoder SAE and SOFTMAX is established and is combined with a pipeline fault type so as to construct a fault diagnosis model, and then the fault classification of the gas pipelines is realized; and by virtue of unsupervised automatic learning characteristic parameters in deep learning and a supervised fine adjustment network, the problems that the characteristic parametersof fault diagnosis of the pipeline are based on experiences and the diagnosis accuracy rate is low are effectively solved, the fault diagnosis of the urban gas pipelines in the working process is relatively rapid, and the stability, precision and reliability of the fault diagnosis are improved.

The invention discloses a multi-classifier-based convolutional neural network quick classification method. According to the method, an activation function and a linear classifier are added after each convolutional layer except the last one. During network training, image features of the convolutional layers are obtained firstly and the classifiers after the convolutional layers are trained by using cross entropy loss functions. After the training, the activation functions are adjusted to enable the classification accuracy to reach the best. During execution of an image classification task, the classifiers of all the layers are activated in sequence in a forward propagation process, the image features after convolution are subjected to calculation analysis through the classifiers, a judgment value is obtained, and if the judgment value meets the activation requirements of the activation functions, classification results of the classifiers are directly output and the classification process is ended. On the contrary, forward propagation activates the next convolutional layer to continue executing the classification task. The method can classify images easy to classify in advance to finish the forward propagation process of a network, so that the network classification speed is increased and the classification time is shortened; and the method has good practical value.

The invention relates to a consistency constraint-based self-learning small sample remote sensing image classification method, which is characterized in that a self-learning method is embedded into adeep convolutional neural network, a self-learning small sample remote sensing image classification method is provided, and the advantages of the self-learning method and the deep convolutional neuralnetwork are comprehensively utilized. In the iterative training process, a consistency discrimination criterion is adopted to continuously generate pseudo label samples, so that label data of small samples are expanded, and the negative influence of false samples of the pseudo labels on the model is reduced by using an adaptive weight l. With the proceeding of the training process, the obtained network classification accuracy is gradually increased, and the problem processing capability of the model is gradually enhanced. Compared with an existing remote sensing image classification method, the method does not depend on a large number of labeled images any more, very high classification accuracy can be obtained under the condition of small samples, and the method has deeper practical significance.

The invention relates to the technical field of image detection and processing, simultaneously relates to the field of bioinformatics, and in particular relates to a method for comprehensively analyzing gene sub-graph similarity probability current by use of multiple image detection technologies. The method comprises the following steps: A, data preparation of a human body gene sequence total graph and a target gene sub-graph; B, detecting the gene sub-graph similarity probability current by use of a CNN (Convolutional Neural Network); C, detecting the gene sub-graph similarity probability current by use of HOG+SVM classification; D, detecting the gene sub-graph similarity probability current by use of Adaboost+LBP feature algorithm; E, detecting the gene sub-graph similarity probability current by use of a standard correlation coefficient template matching method; F, comprehensively analyzing the probability current respectively obtained in the step B, step C, step D and step E by use of a BP neural network classifier to obtain the final probability current after the weighted summation. The method disclosed by the invention can be applied to disease gene detection and capable of fast and accurately detecting whether the human body gene sequence contains the disease susceptibility gene and predicting the disease risk of the body.

The invention discloses a method for detecting defects of a screen printing area of a smart phone glass cover plate based on machine vision. The method comprises the following steps of: collecting animage of a mobile phone screen; reading related parameter information; detecting a window; performing major defect extraction on the outline of the screen printing area of the mobile phone cover plate, dividing a detection area into the screen printing area, a hole and character area, a light band area and an interference area; obtaining the defects of each area; correcting the outline of the mobile phone cover plate to obtain the defect information of edge breakage; performing dotted, linear, and planar defect classification by using a neural network classifier; screening the defects according to the standard of defect definition; performing deep defect classification on the linear defects, IR hole defects, and character defects by using deep learning, wherein the linear defects comprisebroken filaments, scratches, IR hole defects, and the like; and counting the form information of various defects. The method can realize the universal application of various models, performs on-line adjustment according to different detection standards, and can quickly and accurately extracts defects such as pocking marks, broken filaments, scratches and dirt.

The invention discloses a diagnosis method of a rolling bearing fault. The method comprises the following steps of S1, collecting an acceleration signal of a rolling bearing; S2, using a discrete wavelet transform method and a soft threshold method in a combination mode and carrying out de-noising processing on the acceleration signal; S3, carrying out segmentation on a time sequence of the acceleration signal after the de-noising processing and extracting a sample; S4, through more than two self-encoding networks, constructing a stack self-encoding network framework, and extracting characteristic information of the sample; S5, using the characteristic information of the sample to train at least one BP nerve network classifier; and S6, according to a fault diagnosis model established through using known fault data to train the at least one BP nerve network classifier, determining fault information of the rolling bearing. By using the diagnosis method of the rolling bearing fault, accuracy of the fault diagnosis is greatly increased.

The invention belongs to the technical field of signal processing, and particularly relates to an ADS-B-based radiation source individual identification method and device, and the method comprises thesteps of obtaining an ADS-B pulse signal periodically broadcasted by a transponder in a working state, wherein the ADS-B pulse signal comprises a frame header and a data part; extracting individual characteristic parameters of the preamble pulse, decoding the data part to obtain individual identity information, and forming a characteristic vector by the individual characteristics and the individual identity information; dividing the feature vectors into a training data set and a test data set which are used for training a test neural network classifier; and for the target signal, carrying outradiation source individual identification by using the trained neural network classifier. According to the method, the operation complexity is reduced, and the problems of difficulty in individual information verification, low reliability due to pure dependence on simulation signals and the like in individual recognition are effectively solved. By means of the advantages of FPGA parallel computing and software decoding flexibility, the rapid and efficient decoding of the ADS-B messages is achieved, and the method has the important guiding significance for development of the radar radiation source individual recognition technology.