Patents
Literature
27results about How to "Improve classification" patented technology
Efficacy Topic
Property
Owner
Technical Advancement
Application Domain
Technology Topic
Technology Field Word
Patent Country/Region
Patent Type
Patent Status
Application Year
Inventor
Multi-dimensional training method and device of support vector machine
PendingCN114186620AImprove linear separabilityImprove classification and analysis capabilitiesKernel methodsCharacter and pattern recognitionData linesDiscretization
The invention discloses a multi-dimensional training method and device for a support vector machine, electronic equipment and a computer readable storage medium, and the method comprises the steps: carrying out the discretization of a training sample data set, and obtaining a discretized data set, the discretized data set comprises a plurality of different attributes, and each attribute corresponds to a plurality of feature vectors; calculating a classification contribution parameter of each attribute feature vector to obtain a plurality of classification contribution parameters; performing data mapping on the plurality of classification contribution degrees by using a kernel function to obtain a target function; and optimizing and training the objective function by using a gradient descent algorithm to obtain a support vector machine model. According to the method, different dimension data are mapped through the kernel function, the data gain weight can be determined, the linear separable effect of the mapped dimension data can be improved, and then the classification and analysis capability of the SVM can be improved.
Owner:GUANGDONG POWER GRID CO LTD +1
Intelligent intervention system for circadian rhythm disorder insomnia based on APP
InactiveCN119324032Aimprove accuracyImprove timelinessMedical data miningHealth-index calculationCircadian Rhythm DisordersWearable eeg
The invention relates to the technical field of sleep data processing, in particular to an intelligent intervention system for circadian rhythm disorder insomnia based on an APP. The system comprises a user day and night insomnia type marking and dividing module, a day and night insomnia user sleep parameter acquisition module, a day and night insomnia user disorder degree evaluation module and a day and night disorder insomnia user intelligent intervention execution module, and questionnaire survey and user day and night sleep parameter real-time acquisition can be performed through an APP on a wearable wristwatch. Obtaining a user day and night insomnia severity index and a user day and night sleep category measurement score, and performing day and night insomnia type virtual mark division to obtain various day and night insomnia types; the intelligent intervention work of the circadian rhythm disorder insomnia of the user is executed by monitoring the circadian sleep cycles and sleep quality scoring coefficients of the user corresponding to various users in real time and carrying out quantitative calculation of the rhythm disorder of the heart rate cycle and intelligent intervention pushing of the insomnia of the user. According to the invention, a customized insomnia intervention scheme can be provided.
Owner:THE 305TH HOSPITAL OF THE CHINESE PEOPLES LIBERATION ARMY
Assistant communication system for disturbance of consciousness integrating eye movement tracking and electroencephalogram
InactiveCN119440250AReduce the use effectReduce difficultyInput/output for user-computer interactionBiological modelsConsciousness DisordersElectro encephalogram
The invention discloses a disturbance of consciousness auxiliary communication system fusing eye movement tracking and electroencephalogram, comprising a data acquisition device connected with a multi-mode BCI communication system and used for collecting scalp electroencephalogram signals and eye movement signals; the multi-modal BCI communication system can identify the answer (yes / no) of the subject by detecting and analyzing the multi-modal data collected from the communication normal form, so as to realize online binary communication; the multi-modal BCI communication system comprises a multi-modal target identification network, and the multi-modal target identification network specifically comprises a data preprocessing module used for preprocessing and synchronizing different modal data; the double-flow feature extraction module is composed of two independent multi-scale convolutional neural networks which are respectively used for EEG and eye movement data so as to carry out feature extraction on data of different modals; according to the invention, a novel BCI alternating normal form integrating P300 potential and eye movement tracking is designed; the P300 detection and the eye movement tracking detection have irreplaceability and complementarity.
Owner:SOUTH CHINA NORMAL UNIV
Intelligent archive management system based on multi-type analysis
ActiveCN119646278Areduce occupancyeffective classificationSpecial data processing applicationsOther databases clustering/classificationIntelligent managementData mining
The invention belongs to the technical field of archive management, and particularly relates to an intelligent archive management system based on multi-type analysis, which is characterized in that classification term priorities of various archives are determined on the basis of professional term distribution of classified documents in multi-type archives, and key terms in to-be-classified documents are extracted in actual classification operation, so that the classification efficiency is improved; according to the method and the system, document classification is carried out, and the document classification is matched with predefined classification term priorities to realize efficient classification, so that the classification speed can be improved, the occupancy rate of system resources can be effectively reduced, and high-performance operation of the archive management system is ensured; the classification accuracy is verified by carrying out similarity comparison on the newly classified documents and the classified documents in the classified archive types to which the documents belong, errors in initial classification can be found and corrected in time, rapid re-classification is facilitated, and it is ensured that subsequent document retrieval and access are not remarkably affected.
Owner:ANHUI YIBAI INTERNET TECH CO LTD
An improved infrared ship image semantic segmentation method and system of a domain adaptive Deeplab model
ActiveCN118781344BIncrease the average crossover ratioimprove classification
The application discloses an improved infrared ship image semantic segmentation method and system of a domain adaptive Deeplab model, and belongs to the field of visual semantic segmentation. s and corresponding labels Y S ; an infrared ship image dataset X t is obtained, and the infrared ship image dataset X t is divided into a training set X train and a test set X test ; an improved domain adaptive Deeplab model is constructed; the improved domain adaptive Deeplab model is subjected to semantic segmentation by using the visible light ship image dataset X s , the labels Y S and the training set X train , and an image semantic segmentation result is obtained. The average intersection over union of the infrared ship image dataset is improved by modifying the model architecture and the training strategy of the Deeplab, the accuracy of the trained deep convolutional neural network model is higher, and therefore the classification and positioning capability of the network model for ships in infrared images is improved.
Owner:BEIJING UNIV OF TECH
Cultivated land non-agrochemical intelligent monitoring method and system based on multi-temporal SAR and optical image fusion
PendingCN121999441AMake up for the defects of occlusionimprove classificationImage analysisGeometric image transformationSensing dataNerve network
The invention discloses a cultivated land non-agrochemical intelligent monitoring method and system based on multi-temporal SAR and optical image fusion, and belongs to the technical field of remote sensing monitoring and land resource management. The method solves the problems of information loss and misjudgment caused by cloud and mist shielding or noise interference of a single remote sensing data source. According to the scheme, the method comprises the following steps: acquiring and preprocessing a multi-temporal SAR and an optical image sequence; carrying out space registration; constructing a double-branch deep neural network, respectively extracting radar scattering features and spectral texture features, and fusing the radar scattering features and the spectral texture features by adopting a cross attention mechanism; generating a change intensity graph based on a time sequence change detection module; and extracting non-agrochemical suspected pattern spots through threshold segmentation, and aggregating and outputting a final monitoring result. The method is mainly used for realizing the automatic and high-precision monitoring of the non-agrochemical change of the cultivated land.
Owner:GUANGXI ZHUANG AUTONOMOUS REGION NATURAL RESOURCES ECOLOGICAL RESTORATION CENT
Botnet malicious software detection method based on download link
PendingCN119544333AThere will be no interactionreduce couplingSecuring communicationData setAlgorithm
The invention relates to the technical field of data detection, and discloses a Botnet malicious software detection method based on a download link, which comprises the following steps: firstly, analyzing original traffic, capturing a pcap packet in an environment, and extracting IP, port and url information of each frame in the pcap packet; performing data set arrangement on the data of the pcap packet in the environment; training detection is carried out, firstly, an adjacent matrix is constructed, and then parameter initialization is carried out; carrying out matrix operation on the initialized data, then carrying out multi-channel training, carrying out image convolution layer input, and training a recognition model; performing model training optimization on the trained identification model; and applying the trained model to the data based on the download link to detect the botnet malicious software. The detection method is flexible in design and can adapt to constantly changing malicious software attack means and modes. By dynamically adjusting detection strategies and parameters, the method can quickly respond to and adapt to newly appearing threats, and the efficient detection capability is kept.
Owner:NAT COMP NETWORK & INFORMATION SECURITY MANAGEMENT CENT
Multi-level father-child node management method on APP page
PendingCN119474577AAchieve maintenancereduce consumptionWebsite content managementExecution paradigmsComputer networkSoftware development
The invention relates to a multi-level father-child node management method on an APP page, and belongs to the technical field of APP software development. According to the method, each functional page in the APP is used as a routing node, the routing node which reaches each functional page in the APP operation is recorded, and the routing of each functional page is equal to the sum of the routing of the previous skipped functional page and the node of the previous skipped functional page, so that the routing information is transmitted to the next-level page; according to the method, on the basis of runtime runtime characteristics of OBJective-c of an iOS system, maintenance of each node route is realized through a method of holding UIView by HOOK; and meanwhile, by adding classification and attributes to the NSObject, the routing management of each node is realized. The method is based on the characteristics of the iOS development language, follows the language specification of OBJective-c, provides an interface for the system development language, and expands in a pollution-free and coupling-free manner.
Owner:ULTIMATE IOT (HENAN) TECHNOLOGY LTD +1
Image recognition method and system for particle size distribution of raw casting sand
ActiveCN119131379Breduce statistical errorimprove accuracyCharacter and pattern recognitionBiological models
The present invention discloses an image recognition method and system for the particle size distribution of raw sand for casting, the method comprising: obtaining raw sand particle images and preprocessing them; introducing an instance segmentation branch network into a convolutional neural network model to construct an instance segmentation model; performing sand segmentation on the preprocessed raw sand particle images through a pretrained instance segmentation model; calculating the shape characteristics of the sand particles according to the sand segmentation results, and converting the shape characteristics of the sand particles from pixel size to actual size, calculating the equivalent diameter and shape factor of the sand particles; performing mesh classification according to the equivalent diameter and shape factor of the sand particles, counting the number of sand particles of each mesh, and obtaining the recognition result of the particle size distribution of raw sand for casting. The instance segmentation model constructed by the present invention can realize the instance segmentation of the sticky sand particles, reduce the statistical error caused by the sticking of the sand particles, and perform mesh classification and statistics of the particle size distribution of raw sand for casting according to the shape characteristics of the sand particles, thereby improving the accuracy of the calculation.
Owner:HUAZHONG UNIV OF SCI & TECH
Wiring dredging device for high-voltage switch power distribution cabinet
ActiveCN222654566Ureduce areaprevent prolapseBus-bar/wiring layoutsSubstation/switching arrangement casingsControl theoryMechanical engineering
The utility model relates to the technical field of power distribution cabinets, in particular to a high-voltage switch power distribution cabinet wiring dredging device. A high-voltage switch power distribution cabinet wiring dredging device comprises a power distribution cabinet and a plurality of wire harnesses arranged in the power distribution cabinet, and further comprises a binding structure which is located in the power distribution cabinet, arranged on the outer sides of the wire harnesses in a sleeving mode and used for binding the wire harnesses in the power distribution cabinet together, the binding structure comprises a first clamping block, a second clamping block, a connecting arm and a sliding block, and the first clamping block and the second clamping block are rotationally connected to the two sides of the connecting arm through rotating shafts respectively. The wire harness fixing device has the advantages that the wire harness fixing operation process is simplified, the operation efficiency is improved, adjacent binding structures can be adsorbed together by arranging the magnetizer on one side of the clamping block, bound wire harnesses are arranged, the wire harnesses in a power distribution cabinet are placed more neatly, and the wire harness fixing device is suitable for popularization and application. And the classification and management effects of the wire harnesses are improved.
Owner:CHENGDU SUNSHINE ELECTRIC APPLIANCE IND CO LTD
Data table sorting display method and device, equipment and storage medium
ActiveCN117171396BAccurate sortingimprove classificationOther databases indexingOther databases clustering/classificationDatasheetCluster algorithm
The application provides a data table sorting display method and device, equipment and a storage medium. A plurality of data tables, user attributes, user operation data and marking results of different data table importance degrees of a plurality of users are obtained. A preset first number of business hot words matched with the user attributes are obtained from a preset database according to the user attributes of the users. The importance degrees of the data tables in each user group are obtained by accumulating the marking results of each user group, and the users are displayed with the sorted data tables corresponding to the user groups to which the users belong, so that the actual application of each user can be better approached, and more accurate data table sorting can be provided for the users.
Owner:CHINA UNITED NETWORK COMM GRP CO LTD
Feature fingerprint-based agricultural product origin identification traceability system and method
ActiveCN119807816Bcomprehensive descriptionimprove accuracyCommerceManufacturing computing systemsData miningBiology
The present application relates to the technical field of agricultural product origin identification and traceability, and particularly discloses an agricultural product origin identification and traceability system and method based on characteristic fingerprints, which comprises the following steps: collecting agricultural product samples to obtain a target sample set; marking the target sample set with dominant characteristic fingerprints according to the dominant characteristics of the agricultural products, and performing one-class partitioning on the target sample set according to the dominant marking result to obtain a plurality of one-class sample clusters; marking the target sample set with recessive characteristic fingerprints according to the recessive characteristics of the agricultural products, and performing two-class partitioning on the target sample set according to the recessive marking result to obtain a plurality of two-class sample clusters; comparing the one-class partitioning result and the two-class partitioning result to determine whether to start a successive verification mode or a separate verification mode; based on the successive verification mode, obtaining each verification result, and determining whether to output an actual origin traceability result according to the verification results.
Owner:BEIJING SIECAN TECH CO LTD
A mineral identification and classification method based on hyperspectral unmixing technology
ActiveCN118072186Bimprove classificationeasy to identifyClimate change adaptationScene recognitionComputational scienceMixed cell
The present invention belongs to the technical field of hyperspectral image processing. The present invention provides a mineral identification and classification method based on hyperspectral unmixing technology, comprising the following steps: S1, using a hyperspectral imager to obtain hyperspectral images of relevant minerals; S2, constructing a mineral end-member spectral library; S3, establishing a bilinear spectral mixing model based on the hyperspectral images of relevant minerals; S4, using a robust spectral variable bilinear spatial regularization unmixing algorithm to perform spectral unmixing, and accurately estimate the abundance of each mineral contained in the mixed pixel; S5, according to the abundance of the solved minerals, quantify the corresponding proportion of each mineral, so as to achieve accurate identification and classification. This method solves the problems of complex mixed noise interference and spectral variation in hyperspectral images.
Owner:NANCHANG INST OF TECH +1
A method and device for classification and management of nasopharyngoscope images
The embodiment of the specification discloses a kind of for nasopharyngoscope image classification method, comprising: obtaining available image;By image classification model, the target feature map corresponding to available image is obtained, to determine the image classification result of available image;Target feature map includes: carrying out multiple rounds of feature extraction operation;First round of feature extraction operation includes: after available image is subjected to one or more convolution processing, feature map is obtained, the feature extraction result of this round of feature extraction operation is obtained by one or more parallel convolution processing to the feature map obtained;Each non-first round of feature extraction operation includes: the feature extraction result of previous round of feature extraction operation is obtained after one or more convolution processing, feature map is obtained, the feature extraction result of this round of feature extraction operation is obtained by one or more parallel convolution processing to the feature map obtained;Parallel convolution processing is used to extract different range of features;The feature extraction result of last round of feature extraction operation is used as the target feature map corresponding to available image.
Owner:BEIJING JIMAI HEALTH TECHNOLOGY CO LTD +1
Intelligent detection and classification method and system for diseases and insect pests of dendrobium officinale
ActiveCN119600526Aprevent proliferationavoid harmThree-dimensional object recognitionFeature extractionPoint cloud
The invention provides an intelligent detection and classification method and system for dendrobium officinale diseases and insect pests, and relates to the field of data processing, and the method comprises the steps: obtaining a plurality of pieces of sample data; based on the multiple sample data, establishing an optical flow calculation model, a feature extraction model and a pest and disease classification model; obtaining 3D point cloud data of the dendrobium officinale based on a three-dimensional reconstruction technology according to the three-dimensional data of the dendrobium officinale; through an optical flow calculation model, according to the two-dimensional image sequence of the dendrobium officinale, determining optical flow information of the dendrobium officinale; according to the 3D point cloud data and the optical flow information of the dendrobium officinale, whether an optical flow generation area exists in a leaf area of the dendrobium officinale is judged; if yes, extracting image features of the dendrobium officinale from the two-dimensional image sequence of the dendrobium officinale through a feature extraction model; the pest and disease damage type of the dendrobium officinale is determined through the pest and disease damage classification model according to the image features and the optical flow information of the dendrobium officinale, and the method has the advantage of improving the detection and classification efficiency and accuracy of the pest and disease damage of the dendrobium officinale.
Owner:SUZHOU IND PARK SERVICE OUTSOURCING VOCATIONAL COLLEGE (SUZHOU SERVICE OUTSOURCING TALENT TRAINING & TRAINING CENT)
Engine wear prediction and intelligent maintenance strategy maintenance method and device
The present invention provides an engine wear prediction and intelligent maintenance strategy maintenance method and device, which relates to the field of artificial intelligence technology, and processes the operating state parameters of the engine through a pre-built feature extraction model, an autoencoder and a support vector machine. The parameter optimization algorithm of the feature extraction model dynamically adjusts the search strategy according to the network performance feedback, the number of neurons of the autoencoder is dynamically adjusted by evaluating the importance of the input data, the support vector machine uses the whale optimization algorithm to determine the hyperparameters, and the quantum simulated annealing strategy is used to find the global optimal solution, which can more effectively find the optimal model parameters, and can flexibly adjust the model complexity according to the importance of the data, and also help to find the global optimal solution, which helps to achieve more accurate, efficient and reliable engine wear prediction and maintenance strategy formulation.
Owner:SHANDONG ZERO KILOMETER LUBRICATION TECH CO LTD
Infrared laser composite signal target identification and positioning method based on YOLO positioning thought
ActiveCN119646730AAchieve positioningRealize identificationInternal combustion piston enginesBiological modelsData setTarget signal
The invention discloses an infrared laser composite signal target identification and positioning method based on a YOLO positioning thought, and relates to the technical field of signal identification and positioning. According to the method, the positioning thought of a YOLO series model is innovatively applied to the field of one-dimensional data signals, a network structure model is built according to the thought, a sample data set is built, target recognition and positioning of the one-dimensional data signals can be achieved, and experimental results show that the method can improve the recognition accuracy of the composite signals and improve the positioning accuracy of the composite signals. And the position of the target signal appearing in the segment of data can be determined, and target positioning is realized.
Owner:NANJING UNIV OF SCI & TECH
Large outdoor advertisement identification display fault intelligent diagnosis and early warning method
InactiveCN119669856Aimprove classificationImprove fault diagnosis efficiencyKnowledge based modelsThreshold functionNetwork connectivity
The invention provides an intelligent diagnosis and early warning method for a display fault of a large outdoor advertisement identifier, and belongs to an intelligent diagnosis and fault detection technology. The method comprises the following steps: diagnosing a hardware fault, a software fault and a network connection fault; according to the method, by setting a first threshold function f12 and a second threshold function f23 of each parameter in each fault classification, the safety type of the parameter is judged, and the parameter is judged to be one of a safety parameter, a risk parameter and a fault parameter, so that the classification of the display fault of the large outdoor advertisement identifier is refined; according to the method, the pre-warning system of the multi-way tree is arranged, and the hash values corresponding to the parameter safety types of all the fault classifications are written into the pre-warning framework, so that whether risks or faults exist in large outdoor advertisement identification display or not can be directly determined in real time through the root value of the multi-way tree, and the fault diagnosis efficiency and accuracy are greatly improved.
Owner:GUANGZHOU ZHENXIONG DECORATION ENG CO LTD
An indoor positioning method based on multi-sensor of smart phone
ActiveCN116437003Bimprove classificationHigh precisionParticular environment based servicesSubstation equipmentAccelerometerGyroscope
The present invention discloses an indoor positioning method based on a smart phone multi-sensor, comprising the following steps: (1) reading in the accelerometer, gyroscope and magnetometer data collected by the smart phone sensor module; (2) obtaining the initial direction and initial position, and preprocessing the raw data; (3) calculating the step frequency; (4) transmitting the data for human activity state recognition; (5) calculating the step length under different states; and (6) fusing the obtained data to estimate the movement trajectory of pedestrians indoors. The present invention can be fully applied to large places such as shopping malls, teaching buildings, underground parking lots, residential buildings and other indoor places where active signals such as GPS signals cannot be used for positioning to realize indoor positioning and navigation.
Owner:NANJING NORMAL UNIVERSITY
Switch cabinet monitoring device and signal processing method thereof
PendingCN119595090Aaccurate capturecontact stabilitySubsonic/sonic/ultrasonic wave measurementUsing electrical meansComputer hardwareNerve network
The invention discloses a switch cabinet monitoring device and a signal processing method thereof, and the device comprises an extension probe, a socket connector, a mechanical filter, a first probe, a vibration-charge converter, a charge power amplification module, a signal preprocessing module and a signal processing module. The mechanical filter is connected with the first probe through the socket connector, the first probe is connected with the vibration-charge converter, signal acquisition, filtering and classification are finally completed through the signal processing module, signal preprocessing adopts a band-pass filter, fast Fourier transform and normalized denoising filtering, signal processing adopts short-time Fourier transform, and the vibration-charge converter is connected with the vibration-charge converter. And the convolutional neural network extracts depth features, and classification and anomaly detection are carried out through a full connection layer and a Softmax classifier. According to the invention, the signal capture sensitivity and the anti-interference capability are improved through the spring ejector pin type probe and the multilayer composite mechanical filter structure, meanwhile, a cross entropy loss optimization model is supported, and the classification precision and the anomaly detection reliability are improved.
Owner:CHUXIONG POWER SUPPLY BUREAU OF YUNNAN POWER GRID CO LTD
Scraped car disassembling trolley capable of achieving track-changeable conveying
ActiveCN222452209Ueasy to placeEasy to classifyVehicle dissasembly recoveryCarriagesStructural engineeringControl theory
Owner:宜昌华源报废车辆拆解有限公司
Tailing recovery device
The utility model discloses a tailing recovery device which comprises a recovery machine body, the recovery machine body comprises a recovery pool, a feeding port is formed in the recovery pool, and an accidental discharging port is formed in one end, adjacent to the feeding port, in the recovery pool; a power supply is switched on, an external water pipe is connected with a water conveying pipe, then tailings are put into a feeding port, a driving motor drives an electric rotating shaft to rotate, an adsorption magnetic disc is driven to adsorb minerals in the tailings, and the minerals in a groove of the adsorption magnetic disc are scraped out through a scraping plate; minerals are flushed out of the scraper through the water outlet pipe and fall into the screening plate, waste water generated by flushing and the minerals can be conveniently separated through the screening plate, the operation steps of separating the minerals and the waste water by workers in the later period are reduced, the workload of the workers is reduced, and the working efficiency is improved; and meanwhile, different particles can be screened through the two sets of screening plates, and classification of tailings is increased.
Owner:CENT RES INST OF BUILDING & CONSTR CO LTD MCC GRP +2
Method for recommending education content based on knowledge graph and related product
PendingCN119397083Aimprove classificationfully reflectData processing applicationsSpecial data processing applicationsUser inputKnowledge graph
The disclosure discloses a method for recommending educational content based on a knowledge graph and related products, in which the knowledge graph is constructed in a knowledge base containing information and examples corresponding to a plurality of knowledge points, such that a plurality of entities and edges of the knowledge graph represent the plurality of knowledge points and their mutual relationships, respectively. The method comprises the following steps: inputting a recommendation request by a user; identifying a starting entity representing the starting knowledge point in a knowledge graph based on the data; searching one or more additional entities in the knowledge graph along one or more edges starting from the starting entity according to a recommendation policy; and outputting information or examples corresponding to the starting knowledge point and the extended knowledge point in the knowledge base as recommended education content. According to the embodiment of the invention, the recommendation request including the recommendation strategy and the data corresponding to the starting knowledge point is input by the user, so that the user can specify the learning starting point and the recommendation strategy, and various learning requirements of the user are met.
Owner:SHANGHAI PROCESSOR TECH INNOVATION CENT +1
Method and system for automatically identifying mobile phone number
PendingCN119415561Aimprove accuracyeffective classificationDigital data information retrievalSpecial data processing applicationsMobile Telephone NumberPrivacy protection
The invention discloses a method and a system for automatically identifying a mobile phone number. The method comprises the following steps of: establishing a global standard mobile phone number regular expression library; obtaining a plurality of chat record texts, performing encryption transmission and decryption analysis, and obtaining a mobile phone number related character sequence according to the decrypted chat record texts; performing character filtering on the mobile phone number related character sequence to obtain a mobile phone number list; matching the mobile phone number list by using a global standard mobile phone number regular expression library to obtain a mobile phone number; judging whether the mobile phone number is in a number database; if yes, attribution and operator information are stored from a number database; and if not, initiating a query request to the operator through a safe API interface to obtain the attribution and operator information, and updating the attribution and operator information to the number database. The invention relates to the technical field of communication, and solves the technical problems of insufficient accuracy, data security and privacy protection of an automatic identification method in an existing customer service chat system.
Owner:SHENZHEN BADU CLOUD-COMPUTING INFORMATION TECH CO LTD
Cross-data heterogeneity intelligent optimization method, system and device based on data quality dynamic decision
ActiveCN119625499Aimprove classificationImprove segmentationBiological modelsRecognition of medical/anatomical patternsData setReal time analysis
The invention relates to a cross-data heterogeneity intelligent optimization method, system and device based on data quality dynamic decision. The method mainly comprises the steps of S1, data collection and preprocessing, S2, data analysis and imaging quality division are conducted on preprocessed data, a training data set is formed, S3, a model is constructed and trained, the model comprises a backbone network and a plurality of decoders, the backbone network is used for extracting universal features, the decoders are used for processing specific tasks of different quality data, and the decoder is used for processing the specific tasks of different quality data. The training of the backbone network is joint training by uniformly using a plurality of training data sets with different qualities, and the training of the decoder is one-to-one training for the training data set with single quality, and S4, inputting a to-be-optimized data set into a trained model, and analyzing pixel value distribution and characteristic value distribution of input data in real time to obtain a to-be-optimized data set. And automatically matching the most suitable decoder to carry out output optimization. The method is beneficial to stable and accurate segmentation or classification of the image by a subsequent model.
Owner:MACAO POLYTECHNIC INST
Soil chaotic data recognition method based on multi-agent decision
PendingCN122508158AImprove recognition rateImprove unknown identification capabilitiesData setSoil heavy metals
The application relates to the field of agricultural production and environment monitoring technology, and particularly discloses a soil chaotic data identification method based on multi-agent decision, which comprises the following steps: acquiring sample data of soil and constructing a data set; constructing a multi-agent expert pool and a joint representation network, and training the same by using the data set to obtain a soil heavy metal prediction model; wherein the multi-agent expert pool comprises agents of multiple algorithms; the joint representation network comprises a representation encoder, an expert recognizer and a classification discriminator; collecting observable variable data of soil and inputting the same into the soil heavy metal prediction model to obtain a heavy metal concentration prediction value, a confidence, a data type and resampling suggestions; the multi-agent expert pool and the joint representation network can give point estimation, resampling suggestions, sample-level cognitive divergence, knowledge missing estimation and causal sensitivity and other explanatory indexes, and directly support agricultural management decisions such as whether to resample, whether to manually review and the like.
Owner:NANKAI UNIV
A TCT smear automatic detection method
ActiveCN114998264BImprove detection accuracyReduce missed detectionImage enhancementImage analysisCervical cellsCervical cell
This invention discloses an automatic detection method for TCT smears, comprising: constructing DeepTCT; iteratively training DeepTCT: searching for cervical cells detected by DeepTCT in each image of the test set; if the classification score of the cervical cell is greater than a set threshold, retaining the label of the cervical cell; otherwise, discarding the label; adding samples that reach the set threshold to the training set, and using them together with the existing images in the training set for DeepTCT training; applying the trained DeepTCT to the test set, and outputting the classification and localization results of cervical cells; repeating the iterative training a set number of times; and applying the final DeepTCT to cervical cell morphology detection. This invention significantly improves the detection performance of DeepTCT for cervical cell morphology features through multiple iterative training, and the method of this invention has higher mAP and mAR compared with existing methods.
Owner:CHONGQING MEDICAL UNIVERSITY +1