33 results about "N category" patented technology

There are provided an image coding method and apparatus which can properly reduce the computation cost required for rate / distortion optimization processing in image coding at the time of image compression. First of all, the rate / distortion gradient of a frame of interest is classified to m categories. A category information creating unit (101) acquires a threshold λ for a preceding frame from a predictive value storage unit (104) as a predictive value λ′ for the frame of interest when the frame of interest is to be coded, and further finely segments a category in which the predictive value λ′ is included. A code amount is then calculated for each of n categories in which the predictive value λ′ is included. A code sequence forming unit (103) selects a category having a target rate, and searches for a threshold between an upper limit value Tn(x-1) and lower limit value Tn(x) of the category. S at the end of the processing is selected as the threshold λ, and a code sequence is formed by using the threshold λ.

The invention discloses a method for a personalized recommendation of a Web text. The method comprises the following steps: performing feature extraction on a plurality of kinds of Web texts generated before a certain time t so as to obtain a feature matrix E of a Web text set, and then performing cluster so as to obtain n categories; besides, according to a time span hj from the time when a Web text oj in a Web text subset relevant to the behavior of a certain user ui before the time t to the time t, calculating out a preference influence degree dj to the user ui so as to obtain a pair of category number-influence degree cj of the Web text oj and generate a dynamic preference vector of the user ui; if the preference influence degree of the user and the category of the Web text to be recommended is found to be higher than or equal to a threshold value tan, recommending the Web text to be recommended to the user. According to the method disclosed by the invention, the dynamic influence changed with the time lapse of current preference from the historic behavior of the user is considered, and the method is more accurate in recommendation, has dynamic performance, and more conforms to the actual condition.

The invention discloses a fuzzy clustering based image retrieval method. The fuzzy clustering based image retrieval method comprises the following steps of S11, establishing a characteristic value library for images in an image library and numbering the images ; S12, selecting N images with image separation distances larger than a distance threshold value A1 from the image library and conducting first-time classification on the residual images to form N categories of image sets; S13, conducting the step S12 on the image sets with image quantities larger than a quantity threshold value in the N categories of image sets till the image quantity of each image set is smaller than the quantity threshold value and obtaining M representative points; S14, partitioning all images in the image library into the image sets represented by highest-similarity-level representative points according to similarity levels between the images and the M representative points; S15, conducting characteristic valuing on input images to be retrieved, respectively calculating the similarity levels between the input images and all representative points and selecting a plurality of highest-similarity-level representative points to perform retrieval. The fuzzy clustering based image retrieval method narrows a retrieval range on the basis that retrieval efficiency is ensured, and reduces retrieval working amount.

To reduce the computation cost of rate / distortion optimization in image compression, the rate / distortion gradient of a frame of interest is classified to m categories. A category information creating unit acquires a threshold λ for a preceding frame as a predictive value λ′ for the frame of interest that is to be coded, and further finely segments a category in which the predictive value λ′ is included. A code amount is then calculated for each of n categories that include the predictive value λ′. A code sequence forming unit selects a category having a target rate, and searches for a threshold between an upper and a lower limit value of the category. A value (S) at the end of the processing is selected as the threshold λ, and a code sequence is formed by using the threshold λ.

The invention relates to the technical field of computer security and discloses a virus file scanning method and device. The virus file scanning method comprises the steps of: pre-dividing to-be-scanned files acquired from a computer into N categories, and correspondingly dividing a virus scanning process into N stages according to the N categories, wherein N is a natural number greater than 1; and carrying out all the stages of the virus scanning process according to a preset sequence, wherein in each stage, the files with one appointed category is scanned. According to the virus file scanning method and the virus file scanning method device, by dividing the to-be-scanned files into the N categories and correspondingly dividing the virus scanning process into the N stages, the files of the same category are only scanned in each stage; as the categories of the files scanned in each stage are different, the calculated scanning schedule is relatively precise, and the scanning stages can be controlled by a client / server according to the requirements; and therefore, the problems that the scanning schedule is inaccurate and the scanning process can not be controlled by a client / server in the prior art are solved.

The invention provides a wave beam selection method and apparatus, and relates to the technical field of communication. The delay of initial access flow of user equipment can be shortened. The methodcomprises the following steps: selecting a millimeter waves base station for providing service for the user equipment according to the position information of the user equipment; respectively inputting the position information of the user equipment in a plurality of n classifiers corresponding to the millimeter waves base station, and obtaining the wave beam having the highest votes in the sendingwave beams of the millimeter waves base station to serve as a first target wave beam according to judgment results output by the plurality of n classifiers; and sending indication information used for indicating the first target wave beam selected by the user equipment for the millimeter waves base station to the millimeter waves base station. Each n classifier is used for judging the category towhich a wave beam corresponding to the position information of the user equipment belongs in n categories of the n classifier, the n categories correspond to n sending wave beams of the millimeter waves base station, and n is an integer greater than or equal to 2. The wave beam selection method and apparatus provided by the invention are applied to wave beam selection.

A thematic classification method for a digital text file from an encyclopedic database comprising a category graph. A thematic classification model is developed during a learning phase. For each category node, all articles directly linked to the category node is grouped to obtain, for each category node, a “bag of words.” A term-frequency vector characteristic of the category node is determined. At each category node the term-frequency vector, directly connected thereto, with term-frequency vectors of more specific nodes are combined. During the production phase, the term-frequency vector of the digital text file is calculated. N category nodes in the thematic classification model having the closest term-frequency vectors to the term-frequency of the digital text file are selected.

The embodiment of the invention provides a face image recognition method and device, belonging to the technical field of image recognition and aiming to improve the face image recognition rate. The face image recognition method comprises the following steps: matching the characteristics of the first face image acquired through the near-infrared light and the face image in the preset near infrared light registration set and classifying to obtain the categories and the matching scores of the top M candidates with the highest of matching scores in the first group; matching the characteristics of the first face image acquired through the visible light and the face image in the preset visible light registration set and classifying to obtain the categories and the matching scores of the top M candidates with the highest of matching scores in the second group; normalizing the matching scores of the top M candidates with the highest of matching scores in the first group and the matching scores of the top M candidates with the highest of matching scores in the second group respectively; and merging the same categories to obtain the N categories and the scores which corresponds to the N categories, and taking the categories with the highest scores as the recognition results. The face image recognition method and device are mainly used for face image recognition.

The invention discloses a video key frame selecting method. The video key frame selecting method is characterized in that video frames can be divided into n categories by adopting a clustering method, and the categories are called F; a candidate key frame set C can be calculated; video information amount Ii represented by the candidate key frame ci can be calculated; the maximum candidate key frame ci of the video information amount Ii is selected to be the first key frame set S1; according to the formula .... the next key frame set Sk+1 can be calculated in an iterative manner; the video total information amount represented by the key frame set Sk+1 can be determined; and the key frame set Sk+1 can be output. Starting from the diversity of the key frame sets and the maximum information coverage angle, the key frame sets can be quickly and effectively selected from the classified video frames, and therefore the occurrence of the redundancy frames can be well prevented, and the lengths of the key frame sets can be determined automatically; the video key frame selecting method is simple and easy to realize, and the effect is good.

The invention relates to a method for automatically detecting string music trills, in particular to a method for detecting string music in real time during automatic music transcription so as to solve the problems that the trills have great influence on the automatic music transcription, and a traditional automatic music transcription method can not automatically detect the trills in the music during automatic string music transcription. According to the number of common diapason notes of the string music, the trills are classified into N categories, and the N categories of trill modules are trained into a matched object library through an audio recognition method, and audio signals of the music to be detected are inputted, and the characteristics of the audio signals are extracted to obtain a characteristic vector sequence; the average period of the counted trills is used as a length to segment the characteristic vector sequence, each vector sequence segment is recognized through the audio recognition method, and time segments corresponding to the vector sequences of which continuous M or more than M segments are recognized as the same category of trills are detected to be the time segments of the thrills. The invention automatically detects the trills and eliminates the influence of the thrills on the automatic music transcription.

The invention is applicable to the field of computer vision, and provides an image retrieval method and an image retrieval device. The method includes extracting feature points of each image in an image set to create a feature point library; clustering the feature points in the feature point library to acquire N category, wherein the N is pre-defined as an integer larger than 0; extracting feature points of a to-be-retrieved image, predicating category of the to-be-retrieved image according to the feature points of the to-be-retrieved image, wherein the category is one of the N category; counting the occurrence frequency of N-category feature points in the to-be-retrieved image to acquire front M-category feature points with highest occurrence frequency, wherein the M is an integer larger than 0 but smaller than N; acquiring an image set corresponding to the category, counting the occurrence frequency of the M-category feature points in each image of the image set, and summing up the occurrence frequency to acquire values; outputting front L images with highest values, wherein the L is an integer larger than 0. By the image retrieval method and the image retrieval device, image retrieval results can be acquired rapidly and accurately.

The invention provides a streaming media content distribution method based on data feature dimension reduction coding, which comprises the following steps: step 1) extracting on-demand content data ofa server user, and performing historical content analysis to obtain a feature vector; 2) inputting the feature vector in the step 1) into a Word2Vec network, and training a content vector by using animproved Word2Vec model; wherein the improved Word2Vec model is obtained by training an original Word2Vec model and improving label data; 3) carrying out self-encoding training on the content vectorin the step 2) by utilizing a self-encoder model; 4) extracting an Encoder part in the auto-encoder model, and performing dimension reduction encoding on the content vector trained in the step 3) to generate a dimension reduction encoding content vector; and step 5) clustering the dimensionality reduction coding content vectors in the step 4) by using a k-means model, dividing the streaming mediacontent into n categories according to a clustering result, and respectively distributing the n categories to n edge servers of the corresponding content distribution network.

The embodiment of the invention relates to the technical field of deep learning, and discloses a classification model construction method, which comprises the following steps of: obtaining a feature vector of each category in N categories of known images; clustering the N categories according to the feature vectors to form a multi-layer classification hierarchical tree; taking inner nodes of the multi-layer classification hierarchical tree as nodes of the classification model, and training each node to obtain the classification model. According to the classification model construction method, the classification model and the object recognition method provided by the embodiment of the invention, the constructed classification model has a plurality of hierarchies, and the classification performance is relatively high.

The invention provides a distribution warehouse location selection method and device for multiple distribution points. The method comprises the steps of conducting hierarchical cluster analysis, categorizing the distribution points into n categories, calculating transportation cost of the distribution points, using the distribution warehouse location selection device to get a representative point, and calculate the transportation cost of the representative point lj, finally using the distribution warehouse location selection device to select one distribution warehouse L for multiple distribution points. According to the distribution warehouse location selection device for multiple distribution points, a square panel is fixed on a machine frame, a through hole is formed in the panel, two ends of a draw line are used for hanging balance weights and connecting rings. The distribution warehouse location selection method and device for multiple distribution points use a simple and effective distribution warehouse location selection method and device, the method calculation is simple, the operation is convenient and fast.

The embodiments of the invention provide a data processing method and device. The method includes the following steps: obtaining N categories of body physiological index data of multiple detected objects; determining an association model between any two categories of data with different self-management ability intensity, wherein N is an integer greater than 1; obtaining body physiological index data of the current detected object, and inputting the body physiological index data to the association model to obtain the self-management ability intensity prediction result of the detected object; and feeding back a preset monitoring scheme corresponding to the self-management ability intensity to the detected object according to the self-management ability intensity prediction result. The schemeof the invention can reduce the occurrence rate of complications of the detected objects due to low self-management ability more effectively, and thus, can control the deterioration of the conditionof the detected objects and improve the quality of life.

The invention discloses a LDW false and omitted alarm test method based on a convolutional neural network (CNN). The method comprises steps of S1, disposing a camera; S2, setting a maximum lateral distance L, and averagely discretizing the same into n categories; S3, acquiring a real-time image A, inputting the same to a deep CNN model, calculating the actual distance di of a lane line; S4, determining whether a LDW system has false or omitted alarms; and S5, obtaining the misoperation rate of the LDW system. A test system comprises an image acquisition device, an onboard data acquisition mechanism, an analyzer, and an operation processor. The image acquisition device is connected to the analyzer, and the operation processor is connected to the analyzer and the onboard data acquisition mechanism. The method is easy to operate, high in recognition speed and high in recognition precision, applicable to the lanes in various road conditions. The test system can be just provided with the image acquisition device, the onboard data acquisition mechanism, the analyzer and the operation processor in the simplest manner, can fully automatically identify deviations without an extra lane linemark ruler.

The present invention belongs to the computer vision field and relates to a domain adaptive method. The objective of the invention is to solve the problem of domain migration caused by the difference of the distribution of training categories and test categories in object classification. According to the method of the present invention, an expression described in the descriptions of the invention is adopted to indicate a training data set containing n samples; x in the expression, which belongs to R, indicates a vision space; ZS in the expression, which belongs to a (Z1,... ZN) set, contains N categories; an expression U={Xi}<m>i=1 is adopted to express m test samples from M categories, and training categories and test categories do not intersect with each other; each category is characterized by one vector embedded in a category semantic space y which belongs to R<q>, wherein R is a semantic space or a text description space; p and q represent the dimensionalities of the vision space and the category semantic space; according to the basic principle of a cross-modal embedding method, the training data S are adopted to learn a transfer matrix W<*> which belongs to R<p*q>, so that vision samples can be transferred from the vision space X to the category semantic space y; in a test stage, the transfer matrix W<*> in the learning stage is utilized to map test samples x into the category semantic space. The method of the invention is mainly applied to computer vision conditions.

The invention discloses a neural network weight matrix splitting and combining method. The method is applied to target detection with a one-stage network structure. When target detection is carried out, firstly, the type of an object to be detected is determined, training is carried out by collecting picture data, and an original weight matrix is obtained; according to the original N categories, when one or more categories do not need to be removed any more, the neural network weight matrix is splitted, namely one or more categories are split and extracted in the original weight matrix; when one or more categories need to be updated or new categories need to be added, one or more categories in the original weight matrix are extracted for independent training, and then the categories are combined, added and merged through the weight matrix combination. According to the method, free splitting and merging of the target detection model can be realized, the training time is saved, the training steps are simplified, the degree of freedom is relatively high for different scales, and the method has a certain popularization value.

The invention discloses a big-data processing technology, and particularly relates to a method and a system for evaluating an enterprise customer. The method comprises the following steps that obtaining the data information of a customer, wherein the data information comprises multiple categories of data objects; for each the data object of the category which needs to be clustered, setting n types of description, and setting the threshold value [Lambda] of each category; judging whether the obtained n types of description customer data objects exceed a corresponding threshold value [Lambda], if a result shows that the n types of description customer data objects exceed the corresponding threshold value [Lambda], extracting intact information data corresponding to the category in the customer data information, setting n categories of rating rules for the information data, and carrying out evaluation ranking on all customers according to the cluster analysis result of each category of description; if the result shows that the n types of description customer data objects do not exceed the corresponding threshold value [Lambda], putting into a standby customer database; and for the customers of which the levels rank top in the ranking of each category of description, carrying out data extraction, and putting into an available customer database. By use of the method, the customer can be efficiently and objectively analyzed and evaluated.

An apparatus and a method for guiding tool for a user, the apparatus comprising an organiser comprising a number of activity display zones each corresponding to a category of activities; activity tokens representing activities selected in n categories, each activity display zones receiving for display activity tokens representing activities selected in a corresponding category; an a number of score tokens corresponding to a number of points. The user places, on each activity display zone, activity tokens representing activities selected in the corresponding category, the activity display zones displaying their respective activities tokens forming, for the user, a visual representation of target activities and tracking means of activities actually performed, in each category, of points earned and lost by the user, and of benefits to the user, as a result.

The invention is applicable to the field of computer vision, and provides an image retrieval method and an image retrieval device. The method includes extracting feature points of each image in an image set to create a feature point library; clustering the feature points in the feature point library to acquire N category, wherein the N is pre-defined as an integer larger than 0; extracting feature points of a to-be-retrieved image, predicating category of the to-be-retrieved image according to the feature points of the to-be-retrieved image, wherein the category is one of the N category; counting the occurrence frequency of N-category feature points in the to-be-retrieved image to acquire front M-category feature points with highest occurrence frequency, wherein the M is an integer larger than 0 but smaller than N; acquiring an image set corresponding to the category, counting the occurrence frequency of the M-category feature points in each image of the image set, and summing up the occurrence frequency to acquire values; outputting front L images with highest values, wherein the L is an integer larger than 0. By the image retrieval method and the image retrieval device, image retrieval results can be acquired rapidly and accurately.

The present invention provides an identity authentication method and system. The identity authentication method includes: transmitting electromagnetic waves of different frequencies to an object to be identified, and acquiring a plurality of S parameter; filter the S parameter and perform feature extraction, obtain at least one of the amplitude information, frequency information and phase information in the S parameter, and form a test set; use the S that contains n known objects The training sample set of parameter information trains the support vector machine classifier to generate a recognition model that includes n categories, each of which includes all S parameter information of a known object, where n is a positive integer; using the The recognition model recognizes the test set, and if the output result of the recognition model is one of the n categories, the object to be recognized passes the authentication. The authentication and identification of human body identity can be realized quickly and accurately by using the invention.

The invention provides a verification method which includes: by a verification server, receiving a verification request transmitted by a client; by the verification server, generating a verification page according to the verification request, wherein the verification page comprises N category areas and M element images, and N and M are positive integers; by the verification server, receiving correlation, between the N category areas and the M element images, built by a user; by the verification server, verifying according to the correlation. The verification method is simple to operate, high in man-machine interaction performance and high in safety, user experience can be improved effectively, and crack difficulty of verification codes is increased. The invention further provides the verification server.

The invention discloses a video key frame selecting method. The video key frame selecting method is characterized in that video frames can be divided into n categories by adopting a clustering method, and the categories are called F; a candidate key frame set C can be calculated; video information amount Ii represented by the candidate key frame ci can be calculated; the maximum candidate key frame ci of the video information amount Ii is selected to be the first key frame set S1; according to the formula .... the next key frame set Sk+1 can be calculated in an iterative manner; the video total information amount represented by the key frame set Sk+1 can be determined; and the key frame set Sk+1 can be output. Starting from the diversity of the key frame sets and the maximum information coverage angle, the key frame sets can be quickly and effectively selected from the classified video frames, and therefore the occurrence of the redundancy frames can be well prevented, and the lengths of the key frame sets can be determined automatically; the video key frame selecting method is simple and easy to realize, and the effect is good.

The invention discloses a 3D image processing method used for improving the denoising capacity during image rendering. The method comprises the steps of clustering collected 3D point clouds to obtain N categories, wherein N is a positive integer, each category corresponds to a plane, and the 3D point clouds are obtained according to a 3D object; making the value of i an integer from 1 to N and determining the normal vector of the ith plane according to the normal vector of each point in the ith plane to obtain the normal vectors of N planes totally; making the value of i an integer from 1 to N and projecting the points in the ith plane into a plane according to the normal vector of the ith plane, wherein the plane is a detection plane corresponding to the ith plane; obtaining N detection planes; rendering the N detection planes to obtain the 3D image corresponding to the 3D object. The invention further discloses a corresponding electronic device.

The embodiment of the invention provides a face image recognition method and device, belonging to the technical field of image recognition and aiming to improve the face image recognition rate. The face image recognition method comprises the following steps: matching the characteristics of the first face image acquired through the near-infrared light and the face image in the preset near infrared light registration set and classifying to obtain the categories and the matching scores of the top M candidates with the highest of matching scores in the first group; matching the characteristics of the first face image acquired through the visible light and the face image in the preset visible light registration set and classifying to obtain the categories and the matching scores of the top M candidates with the highest of matching scores in the second group; normalizing the matching scores of the top M candidates with the highest of matching scores in the first group and the matching scores of the top M candidates with the highest of matching scores in the second group respectively; and merging the same categories to obtain the N categories and the scores which corresponds to the N categories, and taking the categories with the highest scores as the recognition results. The face image recognition method and device are mainly used for face image recognition.

The invention relates to a method for automatically detecting string music trills, in particular to a method for detecting string music in real time during automatic music transcription so as to solve the problems that the trills have great influence on the automatic music transcription, and a traditional automatic music transcription method can not automatically detect the trills in the music during automatic string music transcription. According to the number of common diapason notes of the string music, the trills are classified into N categories, and the N categories of trill modules are trained into a matched object library through an audio recognition method, and audio signals of the music to be detected are inputted, and the characteristics of the audio signals are extracted to obtain a characteristic vector sequence; the average period of the counted trills is used as a length to segment the characteristic vector sequence, each vector sequence segment is recognized through theaudio recognition method, and time segments corresponding to the vector sequences of which continuous M or more than M segments are recognized as the same category of trills are detected to be the time segments of the thrills. The invention automatically detects the trills and eliminates the influence of the thrills on the automatic music transcription.

The invention relates to a dental jaw tray design method, an apparatus, a user terminal and a storage medium. The method comprises the followings steps: acquiring dental jaw scanning data of patientshaving different identity information; establishing a unified coordinate system for a plurality of the dental jaw scan data; obtaining a dental jaw parameter of each of the dental jaw scanning data inthe coordinate system; dividing the dental jaw scanning data into N categories according to the dental jaw parameter, and determining the shape of the corresponding dental jaw tray according to eachtype of the dental jaw scanning data; wherein N is 2,3,4 or 5. According to the dental jaw tray design method, the apparatus, the user terminal and the storage medium, the dental jaw scanning data ofthe patients having different identity information is obtained, the dental scanning data can be classified according to the identity information of the patients, so that the shape and size of the dental jaw tray are more suitable for Chinese and even Asians, the accuracy of the shape and size of the dental jaw tray can be improved, and the comfortableness of the patient can be increased while a mold is taken.