39 results about "Higher order tensor" patented technology

The present invention discloses a deep neural network compression method based on block item tensor decomposition. The method comprises the steps of: obtaining a framework of a deep neural network, respectively converting a weight matrix W and an input vector x to a high-order tensor (u)W( / u) and a high-order tensor (u)X( / u), performing block item tensor decomposition processing of the high-ordertensor (u)W( / u), replacing a full-connection layer of the deep neural network with a block item tensor layer, and employing a back-propagating algorithm to perform training of the deep neural networkreplaced in the step D. The deep neural network compression method based on the block item tensor decomposition employs a block item tensor decomposition method to construct the block item tensor layer to replace the full- connection layer of the original deep neural network, the features of symmetry and index expression capability of the block item tensor layer are employed to greatly compress the parameter amount of the full-connection layer and maintain the classification precision of the original network.

The invention discloses a tensor-model-based infrared dim target detecting method. The method comprises an off-line training stage and an online detecting stage, wherein the off-line training stage comprises S1, establishing a multi-characteristic sample database and S2, constructing high-order tensor models from the database, and the online detecting stage comprises S3, analyzing whether an unknown image sequence contains dim targets through the tensor models to achieve target detection. By establishing the tensor detection models, the tensor-model-based infrared dim target detecting method avoids the difficulty in identifying targets with complicated backgrounds and dim signal to noise ratios and can well detect targets in different scenarios and with dim signal to noise ratios and improve the detecting accuracy.

The invention discloses a multi-visual angle gait recognition method and a system based on higher-order tensor subspace learning, which belong to the field of intelligent recognition. A gait video is acquired from multiple representational angles, and a gait sequence image is obtained through framing interception; background extraction, background subtraction and binary processing are carried out on the gait sequence image respectively, black and white visual effects are presented, and a contour sequence under the multiple visual angles is obtained; the contour sequence is converted to tensor data; a higher-order discriminant tensor subspace analysis algorithm based on graph embedding obtained after expanding DTSA on the basis of multilinear discriminant analysis and a graph embedding principle is used for carrying out dimension reduction and feature extraction on the tensor data; and according to the extracted and obtained multi-visual angle gait features, the gait features are subjected to similarity measurement, and a recognition result is obtained. The method is simple, the cost is low, person identity authority detection and disguised person identity authentication can be automatically carried out on a particular place, and safety protection on the monitored place and identity authentication in multiple conditions can be effectively improved.

Computer-readable media having computer-executable instructions and apparatuses categorize documents or corpus of documents. A Tensor Space Model (TSM), which models the text by a higher-order tensor, represents a document or a corpus of documents. Supported by techniques of multilinear algebra, TSM provides a framework for analyzing the multifactor structures. TSM is further supported by operations and presented tools, such as the High-Order Singular Value Decomposition (HOSVD) for a reduction of the dimensions of the higher-order tensor. The dimensionally reduced tensor is compared with tensors that represent possible categories. Consequently, a category is selected for the document or corpus of documents. Experimental results on the dataset for 20 Newsgroups suggest that TSM is advantageous to a Vector Space Model (VSM) for text classification.

The invention relates to the technical field of medical instruments, provides a novel practical method for extracting MRI (magnetic resonance imaging) parameters of structural abnormality of white matters of a human brain, and aims to achieve the purpose of objectively evaluate the lesion degree by analyzing anisotropy of tissues through a certain characteristic extraction method. The invention adopts the technical scheme that the method for extracting high-order tensor characteristic parameters of diffusion kurtosis tensor imaging comprises the following steps: a subject collects diffusion weighted signals of tissues along a plurality of directions on a magnetic resonance scanner, then a second-order or fourth-order tensor that reflects diffusion and distribution probability density function of water molecules in the tissues can be obtained through fitting, and lesion characteristics of diffusion and distribution abnormality of the water molecules, caused by structural abnormality of the tissues, can be obtained through certain tensor analysis, wherein parameters from two different angles, including diffusion coefficient and kurtosis coefficient, are mainly included. The invention is mainly applied to design and manufacture of medical instruments.

A method is provided for scalable supervised high-order parametric embedding for big data visualization. The method is performed by a processor and includes receiving feature vectors and class labels. Each feature vector is representative of a respective one of a plurality of high-dimensional data points. The class labels denote classes for the high-dimensional data points. The method further includes multiplying each feature vector by one or more factorized high-order tensors to obtain respective product vectors. The method also includes performing a maximally collapsing metric learning on the product vectors using learned synthetic exemplars and learned high-order filters. The learned high-order filters represent high-order embedding parameters. The method additionally includes performing an output operation to output a set of data that includes (i) interpretable factorized high-order filters, (ii) exemplars representative of the class labels and data separation properties in two-dimensional space, and (iii) a two-dimensional embedding of the high-dimensional data points.

The invention discloses a target tracing algorithm based on a high-order partial least square method. The target tracing algorithm based on the high-order partial least square method comprises an initialization phase, a training phase, a testing phase and an updating phase. A two-dimensional tensor is used for representing image blocks, and internal spatial structure information of the image blocks is stored. Meanwhile, the image blocks of multiple timing sequences and the types of the image blocks are integrated to build a high-order tensor, the relevance between the high-order tensor and the category array of the high-order tensor is analyzed through the partial least square method, and the performance of the target tracing algorithm is largely improved.

Computer-readable media having computer-executable instructions and apparatuses categorize documents or corpus of documents. A Tensor Space Model (TSM), which models the text by a higher-order tensor, represents a document or a corpus of documents. Supported by techniques of multilinear algebra, TSM provides a framework for analyzing the multifactor structures. TSM is further supported by operations and presented tools, such as the High-Order Singular Value Decomposition (HOSVD) for a reduction of the dimensions of the higher-order tensor. The dimensionally reduced tensor is compared with tensors that represent possible categories. Consequently, a category is selected for the document or corpus of documents. Experimental results on the dataset for 20 Newsgroups suggest that TSM is advantageous to a Vector Space Model (VSM) for text classification.

The invention belongs to the technical field of color video recovery, and discloses a color video recovery method and system based on high-order tensor singular value decomposition, and the method comprises the steps that a low-rank part and a sparse part are extracted from a p-order tensor based on a high-dimensional robust principal component analysis (HTRPCA) problem; according to the p-order tensor, conjugate properties of a block diagonal matrix are generated, and an improved algorithm of a tensor product and tensor SVD is proposed; T-SVD is implemented on the tensor, and a new tensor dimension rank is defined; a tensor nuclear norm based on a tensor forward rank is utilized, and a near-end operator and an ADMM algorithm are utilized to solve an HTRPCA problem, so that accurate recovery guarantee of the HTRPCA is realized. Experimental results show that on the premise of theoretical optimal parameters, for generating data and color videos, HTRPCA has superiority in the aspect of processing corresponding problems, and theoretical guarantee is provided for accurate tensor recovery.

The invention discloses a tensor decomposition-based channel state information positioning fingerprint construction method. The method comprises the following steps of: firstly, expressing acquired channel state information (CSI) data as a three-dimensional image; regarding the three-dimensional image as a third-order tensor; then, combining a tensor decomposition algorithm based on a Parallet Factor (PARAFAC) analysis model and an ALS (Alternate Least Squares) iterative algorithm for noise reduction processing of the tensor; then, carrying out single-layer tensor wavelet decomposition on three dimensions of the CSI image by using a tensor wavelet decomposition algorithm, and calculating wavelet coefficients of wavelet sub-components by using an angular second moment; and finally, obtaining the CSI positioning fingerprint corresponding to each reference point coordinate. According to the method, the characteristic that high-order tensors can describe data information and structures isfully utilized, complex data is expressed in a tensor form, noise reduction and feature extraction of tensor images are finally achieved, and the data processing and analyzing capacity is improved.

The invention discloses a millimeter wave system channel feedback method based on tensor parallel compression, and the method comprises the steps: 1, enabling a receiving end to obtain the current downlink channel state information, and then modeling the channel state information into a high-order tensor form; 2, enabling the receiving end to compress the channel state information tensor into a plurality of tensors with the same size by using a tensor parallel compression technology; 3, enabling the receiving end to decompos the compressed tensors through CANDECOMP / PARAFAC (CP) decomposition,and feeding back an obtained factor matrix to the transmitting end through an uplink feedback link; and 4, enabling the transmitting end to recover the factor matrix of the tensor of the original channel state information according to the factor matrix currently fed back to the transmitting end, and reconstructing the original channel state information through the recovered factor matrix. According to the invention, the high-dimensional characteristic of a millimeter wave MIMO-OFDM channel is utilized, and a high-order tensor model is combined with a tensor parallel compression technology, sparse characteristics of a downlink channel are not needed, and the feedback overhead can be effectively reduced under the condition that the feedback precision is ensured.

The invention discloses a similar block accumulation image denoising method and system based on Bayesian tensor decomposition. A plurality of image blocks are divided from an original image; similar image blocks are clustered to form a higher-order tensor; tensor decomposition is performed on the high-order tensor to obtain a joint probability of a tensor matrix and noise precision, a maximum log-likelihood model parameter and a hyper-parameter based on the joint probability are estimated approximately through an alternate iteration method and a denoised high-order tensor is obtained, and finally each image block in the denoised high-order tensor is restored to obtain a denoised low-order image.

The invention relates to the medical imaging technology, in particular to a diffusion tensor imaging method for processing nerve fiber reconstruction in a nerve complex distribution area, device and magnetic resonance equipment. A multivariate high-order tensor model is adopted to describe the diffusion motion in a voxel, and in combination of a high-order tensor decomposition theory, the directions of all fibers in the voxel are solved. The method disclosed in the invention is particularly suitable for reconstructing nerve fibers in the nerve complex distribution area and can effectively process nerve fiber bundle intersection, bifurcation and other conditions. The method, the device and the equipment have important application value in neuroscience, medical imaging and the like.

The invention relates to an intelligent roadside multi-source data fusion method based on Bayesian tensor decomposition. The method comprises the following steps: (1) data preprocessing: collecting traffic flow information at regular intervals through a video and radar detector, marking time labels on lanes, and carrying out the effective recognition and processing of missing and abnormal data to obtain an original data set; and (2) data fusion: constructing a speed tensor, and solving related parameters by using the Bayesian thought to obtain a fused speed tensor. The method is based on intelligent road side heterogeneous data, meanwhile, the Bayesian statistical idea is utilized, the tensor decomposition theory is introduced, the characteristics of high-order tensor mining data characteristics are fully considered, effective fusion of data between different data sources is achieved, and therefore the precision and quality of the data are improved.

Provided is a fiber direction distribution estimating method based on nonnegative higher-order tensor quasi-Newton searching. The method includes the followings steps: reading brain magnetic resonance data, obtaining a magnetic resonance signal S(g) in a gradient applying direction g and a magnetic resonance signal S0 in a gradient non-applying direction as well as gradient direction data, selecting an interested area that is needed, and calculating a diffusion attenuation signal S(g) / S0 in the area; modeling the diffusion attenuation signals S(g) / S0 of all voxels in the interested area one by one to form an ellipsoid distribution model with a diffusion form; and obtaining a diffusion function D(v) by calculating a coefficient vector c of a tensor, then calculating a diffusion function value of each sampling point, fitting the diffusion function values into a diffusion model, and searching an extremum and calculating the fiber direction. A higher-order Cartesian tensor is utilized to achieve fitting of a fiber direction distribution function square root, all fiber direction distribution functions are guaranteed to be nonnegative, the angle resolution is high, and a great experiment effect is achieved.

The invention relates to a color image reconstruction method based on a tensor enhancement technology of a local data block, which comprises the following steps: 1, dividing a to-be-reconstructed image into local data blocks, carrying out structured addressing by taking the local data blocks as tensor enhancement units, converting the to-be-reconstructed image from three-dimensional data into high-dimensional data, and obtaining a high-order tensor; 2 reconstructing the high-order tensor by using a tensor chain nuclear norm minimization model to obtain a reconstructed tensor; and 3 carrying out inverse operation of the step 2 on the reconstructed tensor according to an index position to obtain a converted image, calculating the gray value of each pixel point of the converted image, and restoring the size of the converted image to the original size of the to-be-reconstructed image to complete the reconstruction of the color image. According to the method, a local data block is used as a tensor enhancement unit for structured addressing, the local data block is used as a whole for operation, the complete structure of the local data is reserved, and artifact blocks caused by thorough disruption of the data structure on a reconstructed image are reduced.

The invention relates to the technical field of medical instruments, provides a novel practical method for extracting MRI (magnetic resonance imaging) parameters of structural abnormality of white matters of a human brain, and aims to achieve the purpose of objectively evaluate the lesion degree by analyzing anisotropy of tissues through a certain characteristic extraction method. The invention adopts the technical scheme that the method for extracting high-order tensor characteristic parameters of diffusion kurtosis tensor imaging comprises the following steps: a subject collects diffusion weighted signals of tissues along a plurality of directions on a magnetic resonance scanner, then a second-order or fourth-order tensor that reflects diffusion and distribution probability density function of water molecules in the tissues can be obtained through fitting, and lesion characteristics of diffusion and distribution abnormality of the water molecules, caused by structural abnormality of the tissues, can be obtained through certain tensor analysis, wherein parameters from two different angles, including diffusion coefficient and kurtosis coefficient, are mainly included. The invention is mainly applied to design and manufacture of medical instruments.

The invention provides an MIMO radar two-dimensional direction of arrival estimation method based on constraint tensor decomposition. The method mainly solves the problems that an existing algorithm is high in calculation complexity and poor in estimation precision. According to the method, the high-order tensor model is introduced, the multiple linear relation of MIMO radar multi-pulse receiving data is effectively utilized, and the precision and resolution of direction-of-arrival estimation are improved. The method does not need iteration, is low in calculation complexity and stable in convergence, and can be used for multi-target detection tracking under the condition that the number of space targets is unknown. A high-order tensor model suitable for an MIMO radar with a plurality of transmitting sub-arrays is constructed, the tensor model is reconstructed, and singular value decomposition is performed on an obtained reconstructed matrix, so that a left singular matrix retains a Van der Monde structure constrained by an array structure; according to the special structure of the left singular matrix, phase information of a target coming direction is estimated from a sub-matrix of the left singular matrix by using an ESPRIT-like algorithm, and then two-dimensional direction-of-arrival estimation is realized.

The invention discloses an image restoration method based on smooth Tacker decomposition and high-order tensor Hankering. The method comprises the following steps: 1) inputting a to-be-restored imageto determine a to-be-restored area of the image; 2) constructing a high-order tensor Hankering and discrete total variation model; and 3) constructing a smooth Tacker decomposition and high-order tensor Hankering image restoration model in combination with the step 2), restoring the color image, and finally reconstructing and outputting a high-quality visual data image. The method has the advantages that the image processing efficiency and the image restoration accuracy are both considered.

The invention belongs to the technical field of social network community detection. The invention discloses a method and a device for detecting a large-scale social network community based on high-order tensor decomposition. The detection device based on the high-order tensor decomposition large-scale social network community comprises a network information acquisition module, a central control module, a community search module, a community sharing module, a network compression module, a community analysis module, an influence evaluation module and a display module. According to the method, ablock item tensor layer is constructed by adopting a block item tensor decomposition method through the network compression module to replace a full connection layer in an original social network, andthe characteristics of symmetry and exponential expression ability of the block item tensor layer are utilized, so that the parameter quantity of the full connection layer can be greatly compressed,and the classification precision of the original network can be maintained; meanwhile, the network topology structure can be better utilized through the community analysis module, and the analysis accuracy is high; and distinguishing the types of the core nodes is realized.

The present invention discloses a data cleaning method. The method comprises: acquiring to-be-cleaned data, and according to analysis for noise data distribution in the to-be-cleaned data, obtaining to-be-cleaned fields in the to-be-cleaned data; searching a field which can be subjected to dimension expansion in the to-be-cleaned data, and performing high-order tensor dimension-expanding on the field which can be subjected to dimension expansion to obtain M tensor fieldsets; and carrying out data cleaning on the to-be-cleaned fields by using tensor fields in the tensor fieldsets, which are associated with the to-be-cleaned fields. The present invention also discloses a data cleaning apparatus.

The invention provides a compression method and device of a neural network model. The compression method of the neural network model comprises the following steps: acquiring a parameter matrix in an embedded layer of the neural network model; receiving an order raising instruction, and performing order raising on the parameter matrix according to the order raising instruction to obtain a high-order tensor; decomposing the high-order tensor to obtain a decomposition result of the high-order tensor; and updating the parameter matrix according to the decomposition result to obtain a compressed neural network model. According to the method provided by the invention, on the basis of keeping the performance of the neural network model, the parameter quantity in the embedded layer of the neural network model is effectively reduced, and the volume of the neural network model is remarkably reduced.

The invention discloses a channel state information positioning fingerprint construction method based on tensor decomposition. First, the collected original channel state information (Channel State Information, CSI) data is represented as a three-dimensional image; then the three-dimensional image is regarded as a third-order tensor; then the analysis model based on Parallel Factor (PARAFAC) The tensor decomposition algorithm combined with the Alternate Least Squares (ALS) iterative algorithm is used for tensor noise reduction processing; then, the tensor wavelet decomposition algorithm is used to perform a single-layer tensor on the three dimensions of the CSI image Quantitative wavelet decomposition, and the use of angular second-order moments to calculate the wavelet coefficients of each wavelet subcomponent; finally, the CSI positioning fingerprints corresponding to the coordinates of each reference point are obtained. The present invention makes full use of the characteristics that high-order tensors can describe data information and structures, expresses complex data in the form of tensors, finally realizes noise reduction and feature extraction of tensor images, and improves the ability of data processing and analysis.

The present invention discloses a millimeter wave system channel feedback method based on tensor parallel compression. The terminal uses tensor parallel compression technology to compress the channel state information tensor into several tensors of the same size; 3. The receiver uses CANDECOMP / PARAFAC (CP) decomposition to decompose the above compressed tensors respectively, and obtains The factor matrix is ​​fed back to the transmitting end through the uplink feedback link; 4. The transmitting end restores the factor matrix of the original channel state information tensor according to the factor matrix currently fed back to the transmitting end, and reproduces the factor matrix obtained by restoring To construct the original channel state information, the present invention utilizes the high-dimensional characteristics of the millimeter-wave MIMO-OFDM channel, uses a high-order tensor model combined with tensor parallel compression technology, does not require the sparse characteristics of the downlink channel, and can effectively ensure feedback accuracy. Reduce feedback overhead.

The invention relates to a tensor chain rank-based color image adaptive reconstruction method, and the method comprises the following steps: 1, processing a to-be-reconstructed image to generate a reference image and a to-be-complemented high-order tensor; 2, adding different offsets to the first-order tensor chain rank of the initialized tensor chain rank of the high-order tensor to be complemented, and generating m candidate tensor chain ranks; 3, performing minimization processing on each candidate tensor chain rank to obtain m reconstructed tensors, and converting the reconstructed tensor into a reconstructed image; 4, determining the structural similarity between the reconstructed image and the reference image, and selecting a candidate tensor chain rank corresponding to the reconstructed image with the highest structural similarity as a first-order tensor chain rank; 5, repeatedly executing the operations from the step 2 to the step 4 until the structural similarity between the reconstructed image and the reference image is greater than 0.95, and obtaining an optimal reconstructed image. According to the method, the missing image can be reconstructed without an original data structure.

A processing method for high-order tensor data is disclosed, which can avoid destroying the internal structure of the data during the vectorization process of the image observation sample set, is flexible and simple, can extract the discriminative features of the tensor samples, and simplifies the image observation samples Concentrating a large amount of redundant information in high-order tensor data improves the image processing speed. This processing method for high-order tensor data decomposes high-order tensor data into three parts: shared subspace components, individual subspace components, and noise parts; Quantitative data is expressed as a linear combination of a set of tensor bases and vector coefficients; the variational EM method is used to solve the base tensor and vector coefficients; a classifier is designed to classify the samples to be tested by comparing the marginal distribution of samples.

The invention discloses a multi-attribute volume data compression method based on high-order tensor approximation. It includes preprocessing multi-attribute volume data, decomposing high-order tensors of block data into factor matrices and core tensors, and obtaining approximate high-order tensors according to the reconstruction of factor matrices and core tensors. For each The attribute volume is fused and drawn to complete the multi-attribute volume data compression. The invention greatly reduces the compression rate of the tensor approximation, and at the same time retains the original data of the multi-attribute body, effectively reflecting the target characteristics of the data.

The invention discloses a method for automatically selecting a sparse tensor storage format based on a convolutional neural network. The histogram indicates that the matrix is ​​scaled to a fixed size; 3) The fixed size matrix is ​​used as the input of the convolutional neural network CNN, where the structure of the CNN is designed and customized for the automatic selection of the sparse tensor storage format; 4) The method of supervised learning is used Train CNN and get the trained network model; 5) Take the new sparse tensor as the input of the network model, and get the optimal storage format of the tensor after forward propagation. The invention utilizes the advantages of CNN in the classification problem, and combines the feedforward neural network FFNN to adapt the prediction of the optimal sparse tensor storage format, and on the premise of fully retaining the tensor characteristics, the sparse tensor is effectively converted into a fixed tensor. Matrix input of size, which can be applied to automatic selection of sparse format for higher-order tensors under arbitrary tensor computation.

The invention discloses a sub-array split L-type co-prime array direction-of-arrival estimation method based on fourth-order sampling covariance tensor denoising, which mainly solves the problems that a signal structure is damaged and high-order virtual domain statistics are interfered by noise items in the existing method, and comprises the following implementation steps of: constructing a linear sub-array split L-type co-prime array; modeling a received signal of the L-type co-prime array and deriving a second-order cross-correlation matrix; deriving a fourth-order covariance tensor based on the cross-correlation matrix; fourth-order sampling covariance tensor denoising is realized based on kernel tensor thresholding processing; deriving a fourth-order virtual domain signal based on a denoising sampling covariance tensor; constructing a de-noised structured virtual domain tensor; and obtaining a direction of arrival estimation result through structured virtual domain tensor decomposition. According to the method, high-order tensor statistical distribution characteristics of the constructed sub-array split L-shaped co-prime array are fully utilized, high-precision two-dimensional direction-of-arrival estimation is realized through de-noising virtual domain tensor signal processing, and the method can be used for target positioning.