179results about How to "Comprehensive recognition" patented technology

The invention provides a method and a device for detecting suspicious progresses. The method includes acquiring first characteristic data of the progresses of a browser; matching the first characteristic data in a preset first white list data base; and judging one progress to be a suspicious progress if the first characteristic data of the progress are not in the preset first white list data base, wherein the first characteristic data include Hash values and digital signatures of executable files of the progresses, and the preset first white list data base includes the credible file Hash values and the credible file digital signatures. According to the method and the device for detecting the suspicious progresses, the suspicious progresses can be fully, effectively and accurately distinguished, and safety for users to surf the internet is improved.

The invention provides a method and a system for detecting a vulnerability. The method comprises the following steps of: hooking a function called by a machine code for code execution, acquiring the return address of the function, detecting whether the entry instruction of the function is a normal call instruction or not according to the return address, and if not, determining that a vulnerability is detected. According to the invention, the vulnerability causing a security problem, such as a Web Trojan, can be identified comprehensively and effectively, and the Internet security is improved for a user.

The invention relates to a fast triangle star atlas identification method in astronomy navigation technology field, which includes steps: 1. using star angular distance as a main identification characteristic, using relative magnitude difference of two stars as a auxiliary identification characteristic to determine star atlas identification characteristic quantity, using the star angular distance between two stars, the relative magnitude difference and star number of the two stars composing the star pair as storage basic element, hypothesizing star sensor view field diagonal distance to be d, storing all basic elements with diagonal distance less than or equal to d, constructing a navigation star database and storing the same; 2. extracting observed stars in a observation star atlas, ordering the star from light to dark according to brightness, constructing a observation triangle to be identified; 3. performing triangle identification by utilizing the method; and 4. verifying. The method realizes fast identification of star triangle in the observation star atlas, has advantages of small amount of calculation, stronger fault tolerance capability, faster identification time and navigation database with smaller capacity, and reduces storage pressure of redundant data to hardware system.

The invention relates to the gesture recognition technology in the field of smart TV and intelligent STB(Set Top Box), and discloses an imitated 3D gesture identification system, which solves the problems of high cost, limited identification of the action type and low recognition rate of the gesture in the traditional technology. The system comprises a gesture recognition unit, a main camera and an assistant camera, wherein the main camera is used for collecting the gesture action information in parallel with the surface of the TV terminal screen in front of the TV terminal screen and transmitting to the gesture recognition unit; the assistant camera is used for acquiring gesture information vertical to the face of the TV terminal screen in front of the TV terminal screen and transmitting the gesture information to the gesture recognition unit; the gesture recognition unit is used for performing analysis based on the combination of the gesture action information collected by the main camera and the gesture information collected by the assistant camera, recognizing the gesture action in the three-dimensional space, translating to the corresponding command sending the command to the TV terminal. Furthermore, the invention also provides an imitated 3D gesture recognition method which is suitable for digital TV.

The invention provides a lung focus identification method and device and a realization device. The lung focus identification method comprises steps of obtaining a CT image of a lung, inputting the CTimage into a first identification model and a second identification model which are established in advance, outputting a corresponding first identification result and second identification result, wherein the first identification model is obtained by training a neural network for identifying a focus area having a size greater than a first threshold value and the second identification model is obtained by training the neural network for identifying a focus area having a size smaller than a second threshold value, and determining a final lung focus identification result according to the first identification result and the second identification result. The lung focus identification method and device and the realization device can identify the focus areas of different sizes through different identification models of different neural networks, and can comprehensive identify the areas of big focus and small focus in the CT image while guaranteeing an operation rate.

The invention discloses a network coverage problem identification method and device, and a readable storage medium, which are used for solving the technical problem that the test effectiveness and accuracy of a network coverage problem in the prior art are relatively low. The method comprises the following steps: obtaining MDT sampling data of an MDT sampling point and corresponding signaling data; determining MDT sampling points with missing longitude and latitude information, and performing position backfilling processing on the MDT sampling points with the missing longitude and latitude information according to the corresponding signaling data so as to estimate longitude and latitude information for the MDT sampling points with the missing longitude and latitude information; performingindoor and outdoor scene recognition on a terminal user according to latitude and longitude information included in the MDT sampling data to obtain a scene recognition result; and fitting and displaying the scene identification result and the network quality index information of each MDT sampling point on a rasterized electronic map to present a network coverage problem identification result.

An advertisement filter system comprises a content input interface, a feature analysis module, a decision calculating module, a data recording module, an information base, a command output interface, a manual operation input interface and a machine learning module, wherein the content input interface is used for receiving user generating content from internet interactive products; the feature analysis module is used for analyzing the user generated content, extracting the multiple features of the user generated content and calculating the feature value according to the feature history condition and a manual operation record so as to generate feature vectors; the information base is used for storing various feature data of the user generated content; the decision calculating module is used for comprehensively judging whether the user generated content is filtered or not according to the feature vectors generated by the feature analysis module; the data recording module is used for writing the feature data, categorical data and the manual operation record into the information base; the command output interface is used for arranging a result judged by the decision calculation module into a display / shielding operation command, and synchronizes the display / shielding operation command to the internet interactive product; and the manual operation input interface is used for receiving and analyzing the operation of a filter result modified manually; and the machine learning module utilizes each analysis result and the manual operation record to learn and upgrades the decision calculating module according to the learning.

The invention provides a tristimania diagnosis system and method based on attention and emotion information fusion. The system comprises an emotion stimulating module, an image collecting module, a data transmission module, a data pre-processing module, a data processing module, a feature extraction module and an identification feedback module, wherein the emotion stimulating module is used for setting a plurality of emotion stimulating tasks and providing the emotion stimulating tasks to a subject; the image collecting module is used for collecting eye images and face images of the subject in the emotion stimulating task performing process; the data transmission module is used for obtaining and sending the eye images and the face images; the data pre-processing module is used for pre-processing the eye images and the face images; the data processing module is used for calculating the attention point position and the pupil diameter of the subject; the feature extraction module is used for extracting attention type features and emotion type features; and the identification feedback module is used for performing tristimania diagnosis and identification on the subject. The system and the method have the advantage that the tristimania can be comprehensively, systematically and quantificationally identified by using the attention point center distance features, the attention deviation score features, the emotion zone width and the face expression features.

The invention relates to the field of intelligent teaching, in particular to a student behavior analysis and identification system for an intelligent classroom. The system comprises a coordinate and posture capture module, which obtains the position coordinates of each student and captures the facial orientation posture and hand movement posture of each student. The attention point calculation module obtains the facial orientation posture of the students and calculates the classroom attention points of the students through the facial orientation posture; Behavior recognition module obtains students' attention points and hand gestures, and recognizes students' classroom behavior through students' classroom attention points and hand gestures; Behavior comparison module, to obtain the classroom behavior of each student, and compare with the classroom behavior of its neighboring students, to judge the classroom state of each student; the result feedback module is used to feedback the students' classroom state and feedback it to the teacher in real time. The present invention solves the problem that the prior art is inaccurate in identifying and judging the classroom state of students.

In a mobile device for moving a mobile body in mutually orthogonal two- or three-axis directions, a positional error of the mobile body with respect to a direction orthogonal to preset two-axis directions out of the above two- or three-axis directions is used to determine, by using a sequential two-point method, a straightness error curve representing a change status along one-axis direction out of the above two-axis directions; the straightness error curve determining is repeated for the other one-axis direction out of the above two-axis directions.

The invention provides a video clip tag identification method. The method comprises the steps of obtaining a target video clip; extracting image features and audio features of the target video clip; and analyzing image features and audio features of the target video clip by using a pre-trained multi-label classification model to obtain a tag classification result of the target video clip, the label classification result of the target video clip comprising category labels of the target video clip in at least two dimensions. Based on the scheme provided by the invention, the tag of the target video clip can be comprehensively identified, and the accuracy of a tag identification result can be improved.

The present invention discloses a recognition method for an image object, belonging to basic technology category of the pattern recognition field. The method comprises the following steps of: for an image object as the object and an image object to be recognized, calculating a set of geometric figure characteristic information and retrieval information which reflect main characteristics to form image object characteristic information; directly comparing and calculating the image object characteristic information to be recognized and the image object characteristic information as the object, so as to realize calculation and discrimination for overall similarity of the image objects; realizing quantitative discrimination for local similarity and diversity; and realizing difference discrimination of the rotated image object and the zoomed image object. The method in the present invention has the advantages that: the calculating method of the characteristic information is simple; calculation amount is greatly reduced; and a uniform recognition rule as well as a quick and accurate recognition effect are provided.

The invention relates to a user intention identification method and system. The method comprises a key entity identification step of performing natural language processing technology analysis for userconversation texts by taking words as units to obtain named entities, which serve as user intention parameter candidates, and a user intention judgment step of performing dependency grammar analysison the user conversation texts, performing word-by-word fuzzy matching according to a preset user intention key candidate set to obtain an intention keyword, judging a dependency relationship betweenthe intention keyword and the user intention parameter candidates obtained in the key entity identification step, and outputting a user intention identification result only in the presence of the dependency relationship. According to the method and the system, a user intention can be identified more accurately and comprehensively.

The embodiment of the application provides a media data editing method. The method comprises the following steps: acquiring the physiological data of a plurality of users corresponding to time axes during acquisition of the same media data respectively; counting a time interval formed by corresponding time axes within a specified physiological data range; and editing corresponding media data according to the time interval for serving as feature media data. According to the application, essential segments which are consistent with user preference and interest in the media data can be identified accurately.

The invention discloses a multilayer fund abnormal flow direction monitoring method based on a knowledge graph. The method comprises the following steps: preprocessing transaction flow data within a period of time; extracting an account set involved in the transaction flow; marking different labels on accounts in the set, carrying out aggregation processing on transaction flow data, displaying theaccumulated transaction amount and accumulated transaction times of the main account and the transaction opponent account within a period of time, wherein the knowledge graph operation performance isgreatly improved, and the fund flow direction of the account main body within a period of time is clearly displayed; and finally, importing the processed data into a knowledge graph, and identifyingthe multi-layer fund circulation condition of the account through a graph rule so as to find a suspicious account set; and displaying a suspicious account fund flow direction through a knowledge graphvisualization function. The knowledge graph rule is used for monitoring the abnormal flow direction of the multilayer funds in the field of bank case prevention for the first time, the suspicious account of the abnormal flow direction of the multilayer funds can be effectively recognized, and the visual technology is used for assisting verification and check.

The embodiment of the invention provides a named entity recognition method. The method comprises the steps of obtaining a target text to be subjected to named entity recognition; inputting the targettext into a preset named entity recognition model to obtain a named entity recognition result of the target text, wherein the named entity recognition model is obtained by training a training sample and label information of the training sample; the named entity recognition model comprises a long short-term memory network LSTM layer; the LSTM layer is used for processing each character except the first two characters in the training sample, and the processing process of the LSTM layer comprises the step of carrying out feature extraction on a character vector of the character, an initial feature vector of a previous character of the character and a word vector of the existing word to obtain an initial feature vector of the character if a word exists in the content in front of the characterin the training sample. Compared with the prior art, when the method provided by the embodiment of the invention is applied to perform named entity recognition on the text, the accuracy of the obtained recognition result can be improved.

The invention discloses an intrinsic function recognition method based on a sub-graph isomorphism matching algorithm in decompilation, and belongs to the technical field of decompilation. According to the method, an intrinsic function template library is established, sub-graph isomorphism matching is conducted on intrinsic function templates and target assembling files generated through decompilation on the basis of a control flow diagram, and intrinsic functions which are subjected to compiler optimization and inline expansion in target programs of the target assembling files are positioned. According to the intrinsic function recognition method, inline intrinsic functions in the decomplation process can be recognized automatically, meanwhile, the templates and prototypes of the intrinsic functions are analyzed, the function names, returned values, returned value types and function parameters of the intrinsic functions are recovered, and thus the purpose of promoting the semantics of the inline intrinsic functions is achieved. More type information is provided for type analysis in decompilation through the promoted inline intrinsic functions, the complexity of data flow analysis and control flow analysis is lowered, the level of abstraction of intermediate codes is improved, and the readability of decompilation results is enhanced.

The invention discloses a method for acquiring an alternative name matched pair, comprising the following steps: acquiring contents presenting on each webpage in an internet; extracting a character string pair containing alternative name matched pair in the contents presenting on each webpage according to preset rules; truncating the character string pair containing the alternative name matched pair to acquire the alternative name matched pair. The invention also discloses a device for acquiring the alternative name matched pair. The invention can more and more roundly distinguish possibly emerging alternative name matched pairs, further effectively utilizes the distinguished alternative name matched pairs to improve the experience of users and the utilization rate of data.

The invention discloses a method and a device for identifying a malicious website. The method comprises the steps of obtaining identity information of a to-be-accessed website; matching the identity information with malicious key words in a pre-stored key word bank; in the case of successful matching, identifying the website as the malicious website. No matter whether the malicious website is known or unknown, the malicious website can be identified through the method and the device for identifying the malicious website; the defect that only the known websites can be identified in the prior art is overcome.

The invention provides a multimode identification method and a multimode identification device for a unified Internet of Things identification code. The method comprises the following steps of processing a carrier file when the carrier file is received, and extracting the unified Internet of Things identification code from the carrier file; resolving the unified Internet of Things identification code, and decomposing the unified Internet of Things identification code into more than one piece of single-mode coded information; resolving each piece of single-mode coded information in a matched way to obtain coded contents carried by the single-mode coded information. According to the method, the technical problem of fixed and undiversified identification mode for the coded information of an article in the prior art is solved.

The embodiment of the invention provides a highway safety hazard road section identification method and device. Historical traffic flow data, current traffic flow data and accident influence factors of a road section are acquired, the accident severity of the road section is determined according to the historical traffic flow data and the current traffic flow data of the road section, and the accident frequency of the road section is determined according to the accident influence factors of the road section; then the accident risk value of the road section is determined according to the accident severity of the road section and the accident frequency of the road section; and finally all the road sections are ordered according to the descending order of the accident risk value, and the previously ordered N road sections are determined to be the safety hazard road sections, wherein N is a positive integer. Compared with the static and fixed safety hazard road section determination method, the characteristic that the highway safety characteristics change along with the time-varying characteristics of the accident influence factors can be further reflected so that the safety hazard road sections requiring intensive monitoring in the highway network can be more comprehensively identified, and the identification precision of the safety hazard road sections can be enhanced.

The embodiment of the invention relates to the technical field of communication, and discloses an anti-fraud system and method, a server and a storage medium, and the anti-fraud system comprises a suspicious call analysis module, a network data obtaining module, a fusion analysis module and a warning service module. The suspicious call analysis module is used for identifying the call signaling corresponding to the personal account according to a first preset model and recording suspicious call data; the network data acquisition module is used for discriminating the personal internet surfing data according to a second preset model and acquiring network data for accessing a malicious website; the fusion analysis module is used for performing fraud identification on the suspicious call data and the network data according to a third preset model; and the warning service module is used for intercepting and sending out early warning after judging that the personal account is suspected to becheated. By aggregating the call data and the internet surfing data, the fraud event is more accurately and comprehensively recognized, comprehensive reminding and interception are achieved, and the fraud event prevention and control capacity of family users is effectively improved.

The invention relates to an intelligent seat based on posture recognition, and an intelligent interaction system and method. The intelligent seat comprises an intelligent hardware unit, wherein the intelligent hardware unit comprises at least one pressure sensor, at least one posture sensor, a wireless communication module, a microprocessor module and a power supply module; the pressure sensor is used for detecting pressure borne by the seat and outputting a corresponding pressure sensing signal; the posture sensor is used for sensing the posture change of the seat and outputting corresponding acceleration magnitude and angular velocity magnitude according to the corresponding posture change; the wireless communication module is used for wirelessly communicating with the outside; the microprocessor module is used for receiving corresponding values or signals outputted by the pressure sensor and the posture sensor, processing the values or signals to obtain a seat posture and outputting the same to the outside through the wireless communication module; and the power supply module is used for supplying power to all the modules. The intelligent seat provided by the invention has high recognition precision, simple structure and various functions and can solve the problems that the equipment in the prior art has single function and cannot meet the user requirement.

The embodiment of the invention provides a method and apparatus for identifying abnormal clicking of an advertising space, and relates to the field of computer technology. The method includes the steps of obtaining a clicking matrix according to the obtained clicking log data of the advertising space to be identified, inputting the clicking matrix in a prediction model to obtain the score of the clicking matrix, wherein the prediction model is used to calculate the score of the abnormity of a clicking matrix, and determining the probability that the clicking of the advertising space to be identified is abnormal according to the score of the clicking matrix. The technical problems of inaccurate result due to that the situation of the advertising space itself is not taken into consideration and inability to determine the abnormal clicking behavior can be solved. The determination accuracy is improved, the technical effect of multi-dimensional abnormal clicking behavior identification is achieved. Comprehensive determination of clicking of advertising spaces in different coordinates in a webpage is facilitated.

The invention provides a malicious software identification method and apparatus, and an electronic device. The method comprises the following steps: obtaining to-be-identified software and performingdecompilation on the to-be-identified software to obtain a to-be-processed code; splitting the to-be-processed code into multiple code gene fragments according to a location where a code for calling asystem API is located in the to-be-processed code; detecting the multiple code gene fragments by using a first random forest model to obtain a static detection result; running the to-be-identified software in a sandbox virtual environment to obtain the system API called by the to-be-identified software, detecting the system API called by the to-be-identified software by using a second random forest model, judging whether a malicious behavior exists in a dynamic operation process of the to-be-identified software, and obtaining a dynamic detection result; and judging whether the to-be-identified software is malicious software according to the static detection result and the dynamic detection result. Therefore, the identification accuracy of malicious software can be effectively improved, and malicious software can be more comprehensively identified.

The invention discloses a method and system for a walking robot to cross an obstacle and a robot. The method for the walking robot to cross the obstacle includes detecting whether a first abnormal signal exists by a first detecting device and / or whether a second abnormal signal exists by a second detecting device during the walking of the robot; when only the first detecting device detects the first abnormal signal, detecting, by the first detecting device, a suspected obstacle to acquire first suspected obstacle information, and comparing the first suspected obstacle information with a firstcrossing distance value and / or a first rough surface threshold value in order to determine the type of the suspected obstacle; when the second detecting device detects the second abnormal signal, determining the type of the suspected obstacle. Compared with the prior art, the method for the walking robot to cross the obstacle can detect more obstacle types and responds based on the types by detecting and analyzing the obstacle.

The invention provides a method for identifying high-quality shale by using logging information, and belongs to the field of oil and gas exploitation and physical geography. The method comprises: firstly performing pre-processing on logging information, including performing environment correction, depth correction, and standardization; then performing the following steps: step 1, using a natural gamma-ray logging curve to distinguish a shale bed and a non-shale bed; step 2, using the ratio of a thorium curve and a potassium curve of natural gamma-ray spectrometry log and ratio of the thorium curve and the potassium curve to distinguish sedimentary environment of the shale bed; step 3, if the sedimentary environment distinguished in the step 2 is reduction and low-energy environment, calculating contents of organic matters and clay minerals in the shale bed; step 4, and on the basis of the step 3, identifying high-quality shale beds and non-high-quality shale beds.

The invention relates to the technical field of emotion recognition methods, in particular to an emotion recognition method and system based on nonlinear features. The method comprises a voice collection step: collecting voices of a user; a feature extraction step: processing and analyzing the voice of the user, and extracting the matched features which include audio features, nonlinear features,and semantic features; a model matching step: performing the matching of the matched features, extracted according to the feature extraction step, with a preset emotional models, and finding out the most matched emotion model as a result of emotion recognition. The emotion recognition method and system can comprehensively and accurately analyze and recognize user emotions from multiple angles andaspects according to the voice input.

The invention provides an identification method of a charger for a battery. An information base is previously stored in the main control circuit of the charger. The method comprises the steps: the control circuit arranged in the battery collects the model of the battery and state information, converts the information into a pulse signal of a characteristic frequency, and sends the pulse signal tothe main control circuit of the charger through a circuit switched on when the charger is used for charging the battery; and the main control circuit of the charger receives the pulse signal and identifies the model and battery state information of the battery embedded in the charger referring to the information base according to the frequency of the pulse signal. The invention defines different batteries and battery states through different pulse signals so that the charger more correctly and more comprehensively identifies the battery embedded therein, the charger not only identifies the model of the battery, but also identifies the state information of the battery, thus, the charger automatically adjusts to adapt to the battery in the charging process, so as to prevent damage to the battery caused as the charging parameter and battery parameter are inconsistent in the charging process and prevent the battery from explosion.

The invention relates to an intelligent seat, intelligent interactive system and method for controlling games based on gesture recognition. The intelligent seat comprises an intelligent hardware unit; the intelligent hardware unit comprises at least one gesture sensor for sensing gesture change of a seat and outputting corresponding acceleration values and angular velocity values according to corresponding gesture change; a control board comprising at least two buttons which is used for sending a first game control instruction to the outside; a wireless communication module for communicating wirelessly with the outside; a microprocessor module for receiving corresponding values or signals from a pressure sensor and the gesture sensor and obtaining seat gesture after processing the values and signals and outputting through the wireless communication module; a power supply module for providing electricity for the modules. According to the invention, through adding game control buttons on the seat, through converting gesture into direction control in the game, people can change different gestures in the seat for controlling games so that they can exercise during playing games.