Patents
Literature
Hiro is an intelligent assistant for R&D personnel, combined with Patent DNA, to facilitate innovative research.
Hiro

35results about How to "Reduce misleading" patented technology

Method and system for detection of phenotype genes and analysis of biological information

The invention discloses a method and a system for detection of phenotype genes and analysis of biological information. The method comprises the following steps of: performing local comparison between the sequence of target genes and nucleotide sequences of predicted genome coding regions of all allied species respectively, and obtaining a homologous sequence according to the local comparison results; screening the homologous sequence to obtain the homologous genes; extracting a DNA sequence from the homologous genes, converting the DNA sequence into a protein sequence, performing the overall comparison, and then converting the protein sequence into the DNA sequence; constructing a gene tree for the DNA sequence obtained by conversion, and working out Dn / Ds of each branch; and performing positive selection detection and detecting positive selection loci. The method and the system provided by the invention have the advantages of quickly processing a large amount of data information and discovering more accurate information, reducing misguiding of pseudogenes on analysis of biological information and helping to acquire the biological information to explain more biological problems.
Owner:BGI TECH SOLUTIONS

Speech recognition text processing method and system

ActiveCN106373558AReduce misleadingReach the function of prompting the userSpeech recognitionSpeech identificationSpeech sound
The invention discloses a speech recognition text processing method and a speech recognition text processing system. The method comprises the following steps: acquiring to-be-recognized speech data; conducting speech recognition on the speech data so as to obtain a speech recognition result corresponding to the speech data, wherein the speech recognition result is a recognition text corresponding to the speech data; conducting abnormal speech detection on the speech data so as to determine abnormal speech in the speech data; marking a part, corresponding to the abnormal speech, in the recognition speech; and providing the marked recognition text to a user. With the application of the processing method and the processing system disclosed by the invention, the recognition text of the abnormal speech data can be automatically detected and marked out so as to prompt the user; therefore, misleading to the user is reduced.
Owner:IFLYTEK CO LTD

Car transparent A column device capable of adjusting height to adapt to eye point of driver

The invention discloses a car transparent A column device capable of adjusting the height to adapt to an eye point of a driver. An A column display screen information processing control unit is in communication connection with a camera. The input end of an A column display screen is in communication connection with the A column display screen information processing control unit. A display screen adjusting control HMI device and the A column display screen information processing control unit are in communication connection with a vehicle-mounted ECU through a CAN bus. The camera collects road traffic environment image information in front of a car, and the A column display screen information processing control unit conducts perspective transformation and cutting display processing on the environment image information obtained by the camera; the display screen adjusting control HMI device obtains action information on behalf of the driver intention and sends the action information to thevehicle-mounted ECU, and the vehicle-mounted ECU sends corresponding adjusting instruction to the A column display screen information processing control unit; and the A column display screen information processing control unit adjusts corresponding images displayed by the A column display screen so as to output part of covered image information which conforms to the driver subjective feeling anddisplayed through the A column display screen.
Owner:JILIN UNIV

Device and method for bending crossbeam stress corrosion test

Belonging to the field of corrosion science and protection technologies, the invention discloses a device and method for bending crossbeam stress corrosion test, and mainly studies the stress corrosion behavior of metals (including alloys, also be applicable to other materials). The method includes a loading deformation part and an experimental part. Mainly a groove equipped concave clamp with a predesigned fixed deformation amount and a convex clamp are used in coordination, loading bolts apply force, and retain the force in an experimental environment. A same fixed strain can be applied effectively to different samples to safely, conveniently and efficiently assess the material's stress corrosion susceptibility, including the influence of sample composition, microstructure, stress state or corrosive environment and other factors' changes to stress corrosion susceptibility.
Owner:INST OF METAL RESEARCH - CHINESE ACAD OF SCI

A method and system for discovering causality among social network users combining behavioral sequence and text information

The invention provides a method and system for discovering causality among social network users combining a behavior sequence and text information, which comprises the following steps: S1) data acquisition; S2), preprocessing the data at equal intervals with the smallest time unit; 3) optimize that objective function to find the optimal interval by using the time sequence behavior data; S4) reconstructing the text data in the way of merging the text at the time of merging, and representing the text vectorization; S5) calculating the transfer entropy of the text vector sequences of the two users; 6) prune to obtain a user causality network; 7) store and deriving that us causal network; S8) user causality inquiry and visualization. The invention solves the problem of calculating the transferentropy caused by the sparse user activities; infer the user causality of the social network by using the text data, and the amount of information is more abundant than the pure behavior data; and aninteractive system for inferring, inquiring and deriving the user causality is provided.
Owner:GUANGDONG UNIV OF TECH

Automatic guide method for measuring micro hole

The invention discloses an automatic guide method for measuring a micro hole. A left camera and a right camera are used for shooting a probe and a hole to be measured to acquire a left source figure and a right source figure, and an image A and an image B are generated via edge extraction; and oval detection is separately performed on the image A and the image B to acquire oval outlines of the probe and the hole in each image. Four envelop rectangles are acquired according to the four ovals, the images of the probes and holes in the left source figure and the right source figure are captured according to the four envelop rectangles and amplified to generate four images to perform edge extraction, edge points on the outermost layer are reserved to acquire four oval center coordinates, accordingly, the coordinates of the probes and the centers of the holes to be measured in the left source figure and the right source figure are acquired, and at last, the three-dimensional distance between the probe and the center of the micro hole to be measured is acquired by using a calibrated camera parameter, and a three-coordinate measurement machine is driven to complete guidance. According tothe method provided by the invention, the automatic guidance of the probe to the center of the micro hole is achieved, and the measurement efficiency of the micro hole can be greatly improved.
Owner:合肥庐阳科技创新集团有限公司

Method and device for reverse pushing of information to talents

The invention provides a method and device for the reverse pushing of information to talents, and the method comprises the steps: obtaining a resume database; carrying out the cleaning and gathering so as to standardize the resume data in the resume database; extracting entity relation in the resume data, and constructing a relative network database; searching information in the relative network database according to a received querying sentence, and displaying the searched information in a form of a relative network diagram. Through the technical scheme of the invention, the method achieves afunction that a platform carries out the reverse pushing of information to talents, and brings convenience to the talents to know the objective and comprehensive information of related positions, thereby reducing the misguide phenomena, and improving the interviewing success rate. Meanwhile, the method achieves the display of information in the form of the relative network diagram, so the methodis more visual and enables the information display to be more clear, reduces the time of the talents in reading the information while providing the objective and comprehensive information, and improves the efficiency.
Owner:广州合摩计算机科技有限公司

Self-adaptive hash method and device aiming at incomplete heterogeneous data

The invention discloses a self-adaptive hash method and device aiming at incomplete heterogeneous data. The method comprises the steps that when a heterogeneous data sample is lacked, each of multiple modes is subjected to bipartite graph construction, Markov random walk and probability transfer calculation, and adjacency information of the related lacked sample is estimated out; by studying a self-adaptive diagram, the adjacency information of each of the multiple modes are fused to obtain uniform adjacency information, and according to the uniform adjacency information, a uniform hash code is obtained through a hash technology. The hash method can effectively improve the information search performance of the hash technology, obtains the uniform hash code through study, accordingly improves the searching effect of the system, reduces unnecessary information misleading, and reduces inaccuracy and non-scientificity brought by manual set.
Owner:TSINGHUA UNIV

Non-contact vital sign detection method, non-contact vital sign detection device, storage medium and non-contact vital sign detection system

The invention belongs to the technical field of life information monitoring, solves the technical problem that contact type detection equipment in the prior art cannot complete remote nursing of a special group, and provides a non-contact type vital sign detection method, device, storage medium and system. The method comprises the following steps: acquiring video information of a monitoring area;analyzing the body posture of the target object according to the video information of the target object; carrying out vital sign testing on the supine posture of the body posture or the target objectof the prostrate posture; obtaining abnormal information of the vital sign test, and video information and image information corresponding to the abnormal information; and sending the abnormal information and the image information to the client. The invention further comprises a device, equipment, a storage medium and a system for executing the method. According to the invention, the vital sign information of the abnormal body posture can be quickly and accurately obtained by testing the vital sign of the body posture which may generate the abnormal information, and the monitoring efficiency and accuracy are improved.
Owner:武汉星巡智能科技有限公司

Image recognition method and device, computer equipment and storage medium

The invention relates to an image recognition method and device, computer equipment and a storage medium. The method comprises the following steps: performing attribute identification on a to-be-identified image to obtain attribute information of each part of a to-be-identified target in the to-be-identified image, and classifying the to-be-identified image to obtain a classification result; and determining an attribute identification result corresponding to the to-be-identified target from the attribute information of each part of the to-be-identified target according to the classification result. In the above method, the classification result indicates the real part of the incomplete to-be-identified target contained in the to-be-identified image; wherein the attribute identification result guided and output by the classification result is attribute information of a real part of the to-be-identified target; the method belongs to the attribute information of the visible part in the to-be-identified image, and avoids the attribute prediction of the invisible part of the incomplete to-be-identified target in a conventional attribute identification method, thereby improving the attribute identification accuracy of the image identification method.
Owner:MEGVII BEIJINGTECH CO LTD

Picture processing method and apparatus, as well as terminal

The invention relates to a picture processing method and apparatus, as well as a terminal, and relates to the technical field of picture processing. The picture processing method comprises the following steps: analyzing an picture to obtain picture information, and extracting identification information of a photographing device in the picture information; searching a preset photographing parameter of the photographing device in a preset parameter database according to the identification information; and modifying a parameter corresponding to the picture information according to the preset photographing parameter. According to the picture processing method and apparatus, as well as the terminal, a beautified or modified picture can be adjusted, so that the picture is closer to the state of being just photographed, and the information is displayed as real as possible to meet the needs of a user.
Owner:YULONG COMPUTER TELECOMM SCI (SHENZHEN) CO LTD

Image classification method and device, computer equipment and medium

The embodiment of the invention discloses an image classification method and device, computer equipment and a medium, and belongs to the field of image processing. The image classification method comprises the steps of calling a first classification sub-model in a classification model, and performing classification processing on a target image to obtain a first prediction result of the target image; calling a second classification sub-model in the classification model, and performing classification processing on the target image to obtain a second prediction result of the target image; callinga third classification sub-model in the classification model, performing classification processing on the target image to obtain a third prediction result of the target image, training sample data ofthe third classification sub-model comprising a sample image and a result obtained by statistics according to a plurality of annotation results; and outputting the third prediction result and the classification difficulty of the target image, determining the classification difficulty according to the first prediction result and the second prediction result, indicating the accuracy of the model output result through the classification difficulty, and avoiding misleading of the model to the user.
Owner:TENCENT TECH (SHENZHEN) CO LTD

Transaction evaluation method, device and terminal

The invention discloses a transaction evaluation method and device, and a readable storage medium. The method comprises the steps of A, receiving current scoring data sent by a scoring user side; wherein the current scoring data carries a scoring user identifier and a scored user identifier; b, extracting first historical scoring data of a scoring user and second historical scoring data of a scored user and / or a product; c, calculating the current scoring data and the first historical scoring data to obtain first comprehensive scoring data; calculating the current scoring data and second historical scoring data to obtain second comprehensive scoring data; and step D, storing, outputting and displaying the current scoring data, the first comprehensive scoring data and the second comprehensive scoring data. Not only can the scored user be scored and recorded, but also the score of the scored user can be analyzed, processed, stored and recorded through the first historical scoring data ofthe scored user, and bidirectional datamation analysis and recording of the scored user and the scored user are achieved.
Owner:赵谦

Unsupervised cross-modal hash retrieval method based on noisy label learning

The invention discloses an unsupervised cross-modal hash retrieval method based on noisy tag learning. The method comprises: a multi-modal data acquisition step, a dual multi-modal neural network construction step, a similarity pseudo label with noise extracted based on a pre-trained neural network, an unsupervised cross-modal Hash training step based on noise label learning and a final dual cross-modal Hash retrieval test step. According to the method, two dual Hash model groups are set and relatively clean pseudo labels are mutually fed to each other, so that the learning of the model is misguided by noise pseudo labels as little as possible, the final effect is that the model is converged to a better position, and the performance on a test data set is better than that of other unsupervised cross-modal Hash methods.
Owner:NANJING UNIV

Method for automatically identifying correct answers in community question and answer forum based on artificial intelligence

The invention discloses a method for automatically identifying correct answers in a community question and answer forum based on artificial intelligence. The method comprises the following operation steps: (1) establishing a data set; (2) extracting information features of the text pair by using a deep learning method; (3) extracting other features of questions and answers by using rules, splicingthe features and the features obtained in the step (2) into a feature vector, and the format of the feature vector is [the BERT prediction probability, the similarity of current answers and excellentanswers, the similarity of answers and questions, and the difference of days]; and (4) training a machine learning classification model and predicting a new post. The method provided by the inventioncan quickly and accurately judge the answer possibly being the correct answer under one post, and is time-saving and labor-saving.
Owner:BEIJING UNIV OF POSTS & TELECOMM

Knowledge distillation data enhancement method and system based on generative adversarial network

According to the knowledge distillation data enhancement method and system based on the generative adversarial network provided by the invention, through an easy-to-operate data enhancement mode, the positioning capability and the recognition accuracy of the convolutional network to the industrial data set are improved. According to the method, a generative adversarial network is combined to efficiently fit sample data distribution and an algorithm for improving the fault-tolerant rate of a neural network model in knowledge distillation, non-information noise is not introduced when an enhanced sample is generated, and the robustness of the model to an error tag is improved. According to the method, misleading on the model caused by marking errors of positive and negative samples of the industrial data set is reduced, the representation expression performance of the industrial data set on a semantic segmentation task is improved, and the sample feature learning ability of the model under the condition that feature granularity distinguishing is small is improved; The invention plays an important role in applications such as automobile part detection and manufacturing, railway part positioning and the like.
Owner:WUHAN INSTITUTE OF TECHNOLOGY

Interest point sorting method and device

The embodiment of the invention discloses an interest point sorting method and device, and can determine word weights corresponding to segmented words in keywords when the keywords used for searchinginterest points are obtained, wherein the keywords generally comprise one or more segmented words; wherein the word weight of any segmented word can identify the degree of correlation between the segmented word and the search purpose embodied by the keyword; that is to say, it is possible to obtain a product, wherein the higher the word weight of one segmented word is; wherein the word segmentation is more consistent with the search purpose embodied by the keyword; therefore, in the process of determining the sorting value of the interest points in the search result corresponding to the keyword, the keyword is searched; the word weight can be used as a basis for adjusting the sorting value of the interest points; the sorting value is the basis of the display position sequence when the interest points are displayed; therefore, when the interest points are sorted according to the sorting values determined by the word weights, the preferentially displayed interest points are highly related to the high-word-weight segmented words and more accord with the search purpose embodied by the keywords, the possibility of misleading the user is reduced, and the search experience of the user isimproved.
Owner:BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO LTD

Navigation mark detection system

The invention discloses a navigation mark detection system, which is used for remotely detecting a position and working status of a navigation mark, and solves the problem that navigation marks cannotbe detected in time, and detection results cannot be timely fed back to relevant units. The system includes a terminal detection subsystem and a background remote-control subsystem. The terminal detection subsystem is arranged on the navigation mark, and is used for detecting the position and the working status of the navigation mark. The background remote-control subsystem is arranged in a background, and is used for remotely detecting the subsystem. The background remote-control subsystem includes a navigation mark module, which is used for displaying a navigation mark list; a telemetry module, which is used for detecting a navigation mark by a remote control terminal subsystem; a threshold module, which is used for setting a threshold of fluctuation of the working status of the navigation mark; an alarm module, which is used for issuing an alarm when a detection result exceeds a threshold; and a remote-control module, which is used for setting working parameters of facilities on the navigation mark. Therefore, a relevant unit is enabled to adjust the position of the navigation mark and working conditions of the working facilities on the navigation mark in time, and probabilitythat the navigation mark is lost or a passing ship is misled is lowered.
Owner:深圳市闻迅数码科技有限公司

Substance identification method of ion mobility spectrum (IMS) detection instrument based on ion diagram sequence

The invention discloses a substance identification method of an ion mobility spectrum (IMS) detection instrument based on an ion diagram sequence. The method comprises the following steps: 1. executing preliminary substance identification for one time on a preprocessed ion diagram by utilizing a peak-detection algorithm, and storing substances matched in substance identification at this time; 2. continuously executing step 1 at a time interval of Delta t until times of preliminary substance identification reach preset times M, wherein, M is a natural number which is greater than 2; and 3. eliminating interfering substances from the substances matched in preliminary substance identification for M times. In the invention, for certain multi-peak substances, information such as peak quantity, peak position and the like changes with time, so a false alarm is given if only based on the identification result of the preprocessed ion diagram, which causes misguidance on an inspector using the detection instrument. By utilizing the substance identification method based on the ion diagram sequence, probability of the false alarm can be lowered so as to reduce misguidance on the inspector.
Owner:NUCTECH CO LTD

An automatic guidance method for the measurement of miniature circular holes

The invention discloses an automatic guide method for measuring a micro hole. A left camera and a right camera are used for shooting a probe and a hole to be measured to acquire a left source figure and a right source figure, and an image A and an image B are generated via edge extraction; and oval detection is separately performed on the image A and the image B to acquire oval outlines of the probe and the hole in each image. Four envelop rectangles are acquired according to the four ovals, the images of the probes and holes in the left source figure and the right source figure are captured according to the four envelop rectangles and amplified to generate four images to perform edge extraction, edge points on the outermost layer are reserved to acquire four oval center coordinates, accordingly, the coordinates of the probes and the centers of the holes to be measured in the left source figure and the right source figure are acquired, and at last, the three-dimensional distance between the probe and the center of the micro hole to be measured is acquired by using a calibrated camera parameter, and a three-coordinate measurement machine is driven to complete guidance. According tothe method provided by the invention, the automatic guidance of the probe to the center of the micro hole is achieved, and the measurement efficiency of the micro hole can be greatly improved.
Owner:合肥庐阳科技创新集团有限公司

Substance identification method of ion mobility spectrum (IMS) detection instrument based on ion diagram sequence

The invention discloses a substance identification method of an ion mobility spectrum (IMS) detection instrument based on an ion diagram sequence. The method comprises the following steps: 1. executing preliminary substance identification for one time on a preprocessed ion diagram by utilizing a peak-detection algorithm, and storing substances matched in substance identification at this time; 2. continuously executing step 1 at a time interval of Delta t until times of preliminary substance identification reach preset times M, wherein, M is a natural number which is greater than 2; and 3. eliminating interfering substances from the substances matched in preliminary substance identification for M times. In the invention, for certain multi-peak substances, information such as peak quantity,peak position and the like changes with time, so a false alarm is given if only based on the identification result of the preprocessed ion diagram, which causes misguidance on an inspector using the detection instrument. By utilizing the substance identification method based on the ion diagram sequence, probability of the false alarm can be lowered so as to reduce misguidance on the inspector.
Owner:NUCTECH CO LTD

System resource allocation method and device, and equipment

The embodiment of the invention relates to the technical field of artificial intelligence, and discloses a system resource allocation method and device, and equipment, and the method comprises the steps: obtaining a specified information set with feature data used for representing user risk features, and a tag set; the specified information set at least comprises a prediction sample set and three reference sample sets; the label set comprises risk categories corresponding to the reference samples; correspondingly constructing at least three sub-classifiers by utilizing at least three reference sample sets and the label set; performing risk classification on the prediction sample by using each sub-classifier to obtain a risk classification result of the prediction sample relative to each sub-classifier and a misclassification cost of the prediction sample; and optimizing each sub-classifier according to the prediction sample set, the risk classification result of the prediction sample and the misclassification cost of the prediction sample to obtain each optimized sub-classifier, and allocating system resources to the target user based on the risk prediction result of each optimized sub-classifier for the target user.
Owner:INDUSTRIAL AND COMMERCIAL BANK OF CHINA

Model training method and device, text processing method and device, electronic equipment and medium

The invention provides a model training method, a text processing method and device, electronic equipment and a medium. The model training method comprises the following steps: taking a student model corresponding to an (a-1) th round of iteration as a teacher model corresponding to an a th round of iteration; according to the first training sample corresponding to the a-th iteration and the teacher model corresponding to the a-th iteration, training the basic model corresponding to the a-th iteration to obtain a student model corresponding to the a-th iteration, wherein in the training process, a loss function is determined according to a first probability distribution vector and a second probability distribution vector, the first probability distribution vector is a probability distribution vector obtained by inputting a first training sample corresponding to the a-th iteration into a basic model corresponding to the a-th iteration, the second probability distribution vector is a probability distribution vector obtained by inputting the first training sample corresponding to the a-th round of iteration into the teacher model corresponding to the a-th round of iteration; and outputting a corresponding student model under the condition that the convergence condition is met.
Owner:ZTE CORP +1

Problem replying method and device applied to insurance industry, electronic equipment and medium

PendingCN113010654AImprove experienceReduce knowledge reserve requirementsFinanceText database queryingData miningInformation transfer
The embodiment of the invention discloses a problem replying method and device applied to the insurance industry, electronic equipment and a medium. A specific embodiment of the method comprises the following steps: acquiring a user message; identifying the user message to obtain an identification result; generating question reply information based on the identification result and a preset information database; and transmitting the question reply information to a target device with a display function, and controlling the target device to display the question reply information. The embodiment is applied to the insurance industry, wherein the user message can be recognized and analyzed in combination with the information provided by the user, and then a server is helped to provide more objective reply information meeting actual requirements for the user. According to the invention, wrong guidance caused by human subjective judgment is reduced, and the user experience is improved. Meanwhile, the requirement for knowledge reserve of the server is lowered, and then the manpower and time cost for an enterprise to cultivate the server is lowered.
Owner:北京十一贝科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products