173results about How to "Increase breadth" patented technology

The invention discloses a garbage identification evolutionary learning method, a device, a system and a medium based on depth learning. The method comprsies steps of acquiring garbage sample image data; Preprocessing the garbage sample image data; The pre-processed garbage sample image data is used as the input parameter of the neural network, and is compared with the trained garbage recognition model. According to the comparison result, whether the recognition is successful or not is judged. Feeding back the corresponding information of the identified garbage to the garbage sorting and placing mechanism; The image data of garbage sample which failed to be recognized is recognized again by ResNet algorithm, the image data of garbage sample which is successfully recognized by ResNet algorithm is marked, and the corresponding garbage information is fed back to the garbage sorting and putting mechanism, and the garbage recognition model is updated. The garbage sample image data of garbagesample which failed to be recognized is identified again by ResNet algorithm. The successful garbage sample images identified by ResNet algorithm are transmitted to users or maintainers for marking and updating the garbage identification model. The invention greatly reduces the workload of the maintenance personnel and realizes the accurate classification of a large amount of garbage.

Embodiments of the present invention are directed to compositions and methods for anti-HIV (anti-CD4 binding site) antibodies having improved potency and breadth.

The invention discloses a multi-angle drop-hammer impact testing device. The multi-angle drop-hammer impact testing device comprises a test bench for fixing a test specimen, a drop-hammer assembly and a lifting mechanism, wherein the drop-hammer assembly is correspondingly arranged above the test bench by the lifting mechanism; the test bench comprises a test bench main body and an angle adjustment mechanism for adjusting the rotation of the test bench main body, and the drop-hammer assembly is kept to correspond to the test specimen on the test bench main body by a hammer adjustment mechanism. The invention also discloses the test bench of the multi-angle drop-hammer impact testing device using the structure. The test bench main body rotates by a preset angle by the angle adjustment mechanism, so as to simulate the impacts from different angles; the drop-hammer assembly is allowed to impact the test specimen from different angles, and the impacts comprise front-surface impact, the offset impact and the lateral impact, so that the material mechanical properties of the test specimen can be analyzed from different impact angles, and the flexibility and universality of the drop-hammer impact test are greatly improved.

A system and method for performing a search and retrieval of documents in a computer network is presented, wherein the user can conduct a multi-lingual search and receive results in his or her natural language. The system includes steps wherein a user inputs a query in a source language, and selects one or more target languages. The query is then translated into the target languages and a contextual search is performed using the original and translated queries. Once search results are obtained, a language translator utility then identifies the language of the search result and that result is then properly translated into the language of the user. This system is particularly useful for searches over the Internet.

A prediction processing system, method, and method for doing business is disclosed. The prediction processing system can collect, process, and publish event-outcome information. The prediction processing system can dynamically filter participants into groupings and iteratively optimize odds calculations over time. A proposal framework consisting of a various abstract proposal types can model and settle propositions. Game propositions can be automatically generated based on event categorization relationships, and users can themselves automatically establish their own game proposition and programs. The prediction processing system can provide game players or others with access to collective intelligence, including prediction information and information derived from prediction information. The prediction processing system can create groupings of better-performing predictors and provide them with information not generally available or additional stakes to increase the weight of their predictions in odds calculations. The prediction processing system can provide access to event-outcome information on a for-fee subscription basis.

The invention discloses an unmanned aerial vehicles searching-based method and system. First, registration or residence information in personal mobile terminals of persons in distress is searched by using a personal wireless network simulated by unmanned aerial vehicles, information searched by different unmanned aerial vehicles is interacted through an unmanned aerial vehicle self-organizing network formed by communication modules of the vehicles, then information of identity and geographical positions of persons in distress is calculated and determined according to information of persons in distress acquired by unmanned aerial vehicles independently, and finally, combined information is returned to a ground station of a rescue center through a long-distance return module of a return unmanned aerial vehicle. According to the method and system, multiple different communication modules of the unmanned aerial vehicles are used to search and rescue persons in distress in a large range accurately and efficiently.

The invention discloses an accompanying relationship acquisition method and system, equipment and a storage medium. The invention discloses a space-time trajectory-based adjoint relationship acquisition method and system. Identity features with space-time similarity corresponding to different individuals are acquired by excluding identity feature pairs with space-time similarity corresponding to the same individual, first adjoint relationship information is acquired by utilizing the identity features, and only data with space-time similarity and different individuals is calculated. Another aspect, The invention discloses an adjoint relationship obtaining method and system. A second adjoint relationship information among the virtual social identities is acquired according to the virtual social identities corresponding to the mac addresses, and the obtained first adjoint relationship information is utilized, fusion of the first adjoint relationship information and the second adjoint relationship information is realized according to the mac address, and the combined adjoint relationship information is obtained, so that the problem of data sparsity in the original method is effectivelysolved, and the breadth of identity characteristics in the adjoint relationship is further improved.

The invention provides a system and a method for determining an IP address of network equipment using network address translation, which are applied to a system comprising a core network and a private network, wherein when the core network is to determine the IP address of the network equipment in the private network, the method comprises the following steps that: the core network sends an order for acquiring a public IP address of the private network and port mapping information of the network equipment to the private network, and the private network sends the public IP address of the private network and the port mapping information of each network equipment to the core network after receiving the order; the core network sends a message for acquiring a private IP address of one of the network equipment to the private network according to the acquired port mapping information of the network equipment, and the private network sends the to-be-acquired private IP address of the network equipment to the core network after receiving the message for acquiring the private IP address of the network equipment; and the core network sends a notification message containing the public IP address of the private network and the acquired private IP address of the network equipment to the private network.

The invention relates to the field of pilotless helicopters, especially to a modularized pilotless helicopter for task loads. The total design of the pilotless helicopter is carried out by taking task loads as a constituent part of the pilotless helicopter. The pilotless helicopter is divided into an upper part flight platform module (1) and a lower part task load module (2); wherein the flight platform module (1) comprises regular basic components of a pilotless helicopter, the regular basic components consisting of a fuselage (101), a rotor system (102), a flight control system (103), a power system (104), a electric power system (105), cables and interfaces (106); the task load module (2) mainly consists of a load cabin (201); and the two modules are rigorous in anastomosis and firm in connection and form an integral pilotless helicopter. The problems that a conventional pilotless helicopter with a limit capacity in a singularity shape cannot be suitable for various kinds of load devices in various sizes and shapes is solved by the modularized pilotless helicopter for task loads of the invention, and the flexibility of the task load integrated usage and the multitask execution capability of the pilotless helicopter are both improved.

The invention discloses an age estimation method based on deep learning. The age estimation method based on deep learning comprises steps of (1) constructing an age database, (2) performing pre-processing on images of the constructed age database, (3) performing unification and normalization on sizes of aligned images to obtain images with a size being 64*64, (4) taking the obtained images and corresponding labels as inputs of a deep model, using a CNN convolution deep network to train an age estimation model, (5) inputting an age estimation model into a tested image to obtain similarity values of tested images on various kinds of labels, (6) multiplying the obtained corresponding label with the obtained similarity value and then adding obtained corresponding label and the obtained similarity value to obtain a final age estimation result. The age estimation method based on deep learning can obtain a smaller deep model and is fast in operation time and high in an age estimation identification rate. The database comprises massive samples of children and senior people and can effectively identify age of a special group.

The invention discloses a layer-by-layer foundation consolidating method based on a barrel-type foundation negative pressure technology, comprising the following steps: (1) a reserved negative pressure hole is arranged on a cover plate of the barrel top of the barrel-type foundation, a plurality of steel tubes with sieving holes at the side walls are fixedly connected below the cover plate of thebarrel top of the barrel-type foundation, the length of the plurality of steel tubes is lengthened gradually according to the set depth of layered consolidated foundation soil layer in the direction starting from the negative pressure hole to the barrel wall of the barrel-type foundation, each steel tube is parallel to the central line of the barrel-type foundation, and geotextile is laid below the cover plate of the barrel top of the barrel-type foundation; (2) according to set injection negative pressure value, the barrel-type foundation is sunk and positioned by negative pressure; and (3) negative pressure equipment is used for pumping water and/or air by the reserved negative pressure hole and carrying out negative pressure foundation consolidation treatment according to set prepressing value of the foundation soil till the foundation soil is overconsolidated. The layer-by-layer foundation consolidating method realizes layer-by-layer foundation consolidation, and effectively increases the depth and width of the foundation consolidation, thus improving the bearing capability and stability of the barrel-type foundation.

The invention discloses a multi-objective-based building scheme stage energy-saving optimization design mode construction method, and belongs to the field of building design. The method sequentially comprises the steps of integrating a building body and environment information, determining an optimization target and a design variable, establishing a building multi-target prediction model, coupling the prediction model and executing operation, and constructing an energy-saving optimization design mode. According to the method, the building energy consumption, the thermal comfort index and the initial investment cost are considered at the same time in the scheme stage, the support vector machine algorithm in machine learning is adopted to construct the multi-target prediction model, compared with software simulation, target prediction can be converted into a numerical calculation process from a performance simulation process, the prediction efficiency of the performance index is improved, And the synchronism of multi-target prediction can be ensured. The algorithm is specially proposed for the small sample problem, so that a large number of redundant samples can be eliminated, the optimal model is obtained under the condition of limited samples, and the construction efficiency of the prediction model is improved.

The invention discloses a 3D face recognition method, a terminal device, and a readable storage medium. The method comprises: collecting a to-be-verified current face image by a terminal device; according to the current face image and a pre-stored face photo in a face feature library, generating a current 3D face; comparing the current 3D face with the pre-stored 3D face in the face feature library and realizing recognition of the to-be-verified face based on the comparison result. According to the invention, face modeling is realized only by shooting two photos; the matching data are enriched; when a to-be-verified user shoots the own face photo at any angle from 0 to 45 degrees, recognition is done successfully, so that the breadth and accuracy of face recognition are improved.

The invention relates to a json based complex web page component self-defining method and device.The method includes the following steps that 1, atomic-scale page elements on a web page are identified and acquired; 2, atomic-scale components are formed according to the atomic-scale page elements; 3, data models of the atomic-scale components are defined by adopting a json data structure, and functional logics of the atomic-scale components are achieved by adopting javascript; 4, a user input interface is generated according to the atomic-scale components defined in the step 3; 5, the user input interface receives input data, legal verification is conducted on the received input data, and user self-defined service components are formed according to the input data passing the verification.Compared with the prior art, the json based complex web page component self-defining method can achieve description, definition and resolution of the elements of any complex web page, provide the capability of generating component content data interfaces and facilitate component content data input performed by a user.

The invention discloses an AI customer service system based on an official account. The AI customer service system comprises an official account management module, a customer interaction module, a wireless data sending module and a commission computing module, wherein the official account management module is used for customer service management, problem classification management, word segmentation management, fast reply management, keyword automatic reply management, common problem fast reply configuration and the like of the official account, so that the background configuration efficiency is improved to the largest extent; the customer interaction module is used for facilitating timely communication between a staff customer service and a customer while realizing an online real-time chatting function, and sending expressions, files, links and the like to the customer; a staff service module is used for judging an on-line state of the customer service, user access is realized only through the on-line customer service system, so that the situation that a user is not replied all the time after access to the customer service is avoided, and meanwhile, the customer service can also set a star target for the special user, and the user can evaluate the customer service after the problem is solved.

The invention discloses a sale platform based on a smart phone platform. The sale platform comprises a smart phone with an Android or IOS system, a database server and a website server. A merchant service side is installed in the smart phone and can upload related product information in a multithreading mode through a network communication protocol. The database server can store the product information uploaded by the merchant service side, associates the product information with websites set by the website server, and meanwhile transmits purchase information to the merchant service side. The website server can simultaneously share the product information uploaded by the merchant service side to a plurality of domestic and foreign social websites in a one-key mode for transactions. According to the sale platform, sellers can use mobile terminals to take pictures and edit information at any time and any place and meanwhile share the information on domestic and foreign social media in a one-key mode for product sale.

The invention discloses an internet public opinion monitoring method and system. The method includes the steps of collecting texts to be analyzed; detecting the languages which the texts to be analyzed belong to and extracting characteristic words of the texts to be analyzed; inquiring a translation dictionary including languages from the languages which the texts to be analyzed belong to and a reference language according to the characteristic words of the text to be analyzed to obtain translations of the characteristic words of the texts to be analyzed in the reference language; calculating similarity between translations of reference language hot spot public opinions and the translations of the characteristic words of the texts to be analyzed in the reference language, and determining relevancy of the texts to be analyzed and the reference language hot spot public opinions according to the calculation result so that internet public opinion information of other languages related to the reference language hot spot public opinions can be obtained, the internet public opinion monitoring range cannot be restricted to the reference language range any more, and validity and universality of internet public opinion monitoring can be improved.

The invention discloses a virtual identity knowledge graph establishing method for aiming at network thread discovery, wherein the method belongs to the field of network virtual identity tracking. The method aims to settle a technical problem of reducing time in searching virtual identity information for improving network tracking and tracing efficiency. The method comprises the steps of (1), searching the virtual identity information of a user through a plurality of ways which comprise searching user registration information of different Internet websites, directionally crawling user individual information which is disclosed in a preset website, and crawling the virtual identity information of the user based on a searching engine; (2), combining a plurality of sets of user registration information in a user registration information table for generating a data sheet I, (3), combining the data sheet I with a user individual information sheet for generating a data sheet II; and (4), combining a virtual identity information sheet with the data sheet II for generating the virtual identity knowledge graph. The identity information is searched through the virtual identity knowledge graph, thereby reducing time in searching the virtual identity information and improving efficiency in network tracking and tracing.

The invention discloses a commodity recommendation method and device, relates to the technical field of data analysis, and aims to solve the problem of low recommendation accuracy in the prior art. The method mainly comprises the following steps: acquiring recommendation data, wherein the recommendation data comprises user information, commodity data information and commodity text information, theuser information refers to the age, name, registration duration and registration address of a user, the commodity data information refers to digital description information of commodities browsed bythe user, and the commodity text information refers to text description information of commodities browsed by the user; encoding the commodity text information according to a self-encoding algorithm to generate a text feature code; inputting the user information, the commodity data information and the character feature codes into a parallel recurrent neural network, and calculating the probabilitythat each commodity in a commodity library is clicked by a user; and according to the descending order of the probabilities, selecting and displaying a preset number of recommended commodities corresponding to the probabilities. The commodity recommendation method and device are mainly applied to the commodity recommendation process.

The invention discloses a traffic signal lamp control method, device and system, and the method comprises the steps: detecting whether the vehicle passing condition of a first detection region and/ora second detection region of a target lane meets a preset condition or not after the initial green light stage of a signal lamp of the target lane is finished; wherein the target lane comprises a detection blind area, a first detection area and a second detection area which are sequentially arranged from a vehicle line in the direction opposite to the vehicle driving direction, and the vehicle passing conditions of the first detection area and the second detection area are detected through arranged detection equipment; if the preset condition is met, controlling the signal lamp to enter a green flash countdown stage; and if the preset condition is not met, controlling the signal lamp to enter a green light prolonging stage. Through the above mode, the detection precision of the traffic condition can be improved, maintenance is convenient, the green light duration of the traffic signal lamp can be dynamically adjusted in combination with the traffic condition, and traffic jam is avoided.

The invention provides a special-shaped sanding machine. The sanding machine comprises a portal frame, a whole height adjusting frame which is arranged on the portal frame and can vertically move and a grinding device arranged on the top of the whole height adjusting frame; the grinding device comprises a first grinding mechanism, a second grinding mechanism, a third grinding mechanism and a fourth grinding mechanism, wherein the third grinding mechanism and the fourth grinding mechanism are the same in structure; and according to the machine, the height of equipment can be adjusted according to different sample plates, use university of the equipment is improved, the grinding height difference is large, and the maximum height difference can reach 3 cm. Priming paint and substrates can be ground at the same time, and abrasive paper does not need to be replaced. Meanwhile, design is achieved on the basis of the human engineering principle, multi-direction grinding is achieved, and finally dead-zone-free grinding is achieved.

The invention relates to a remote control patrol system based on a metering method for carbon emission in a district. The remote control patrol system comprises a body, an atmosphere detection unit, a lift sampling unit, an air bag unit, a communication display unit, a remote controller, a camera, an obstacle detector, two steering wheels and a power unit, wherein a sampler of the lift sampling unit is arranged on a lift device and is capable of sampling at any height lower than a maximum lift height, a sample transfer channel connects the sampler to an inlet of the atmosphere detection unit, the air bag unit opens when a liquid detector arranged at the bottom of the air bag unit detects a liquid substance, the power unit stops working when the atmosphere detection unit begins sampling detection, the power unit resumes working after detection is finished, and the camera acquires an obstacle image and carries out gray scale identification on the obstacle image when the obstacle detector detects an obstacle.

The invention relates to a XSS (Cross Site Scripting) vulnerability detection method and device, and belongs to the technical field of network security. The XSS vulnerability detection method comprises the steps of: using an initial address and a derived address of a target site as to-be-accessed addresses; loading the to-be-accessed addresses in a browser simulation environment, and acquiring first detection pages; triggering target element nodes in the first detection pages, which have an interaction attribute, by simulating a browser behavior, and obtaining second detection pages; acquiringinjection points in each first detection page and each second detection page; and detecting whether each injection point has a XSS vulnerability. According to the invention, misinformation and reportomission of the XSS vulnerability can be reduced.

The invention discloses a Serverless framework-based multi-engine virus scanning system and multi-engine virus scanning method. The multi-engine virus scanning system comprises a mobile terminal and aServerless cloud network; the mobile terminal comprises a client; the Serverless cloud network comprises a virus engine router and a Serverless scanning cluster; the client is used for uploading to-be-detected files; the virus engine router is used for receiving the to-be-detected files and distributing the to-be-detected files to containers in the Serverless scanning cluster in sequence according to a preset task distribution algorithm; the containers are used for performing virus scanning on the received to-be-detected files by utilizing the corresponding virus scanning engines and sendingscanning results to the virus engine router; and the virus engine router is further used for summarizing the scanning results and generating corresponding final scanning result information. Accordingto the technical scheme, the consumption of calculation resources and power resources of the mobile terminal can be greatly reduced; and the virus searching and killing range and accuracy and the novel virus response speed are improved and increased.

The invention belongs to the technical field of application of chitin and relates to a nano chitin hydrogel and aerogel and a preparation method and application thereof. The preparation method of thenano chitin hydrogel comprises the following steps: (a) pre-treating the chitin raw material to obtain nano chitin dispersion liquid; (b) mixing the nano chitin dispersion liquid with a cross-linkingagent to obtain compound dispersion liquid; and (c) freezing and thawing the compound dispersion liquid to obtain the nano chitin hydrogel. The invention innovatively adopts the freezing and thawing treatment to successfully realize chemical cross-linking of nano chitin; the product has the characteristics of low grade, high strength, high elasticity, high porosity and the like and also has the advantages of mild conditions, strong operability and the like; and a feasible method is provided for preparation of nano chitin gels and has industrialized application prospects.

The invention provides a multi-redundancy unmanned aerial vehicle data transmission system and method. An onboard data redundancy fusion computer obtains flight telemetry data and video image data inreal time; data are selected according to a link data selection mechanism and are packaged; the data are transmitted to a ground data redundancy fusion computer via a corresponding data link channel;and the ground data redundancy fusion computer receives data packets from the data link channels in real time and obtains a data type to which data contents belong according to a data type identifier,verifies the correctness of the data via a verification bit and discards the data of the same data type which arrive through the other data link channels according to a data sequence identifier. According to the multi-redundancy unmanned aerial vehicle data transmission system and method, transmission data contents are selectively selected corresponding to different data link characteristics through the link data selection mechanism, thereby improving the data transmission stability and reliability under the condition of meeting the data transmission content requirements maximally, and improving the information transmission breadth and speed.

The invention belongs to the field of large touch screen applications and particularly relates to a touch table with a square table face. The touch table is suitable for electronic games with multiple people sitting around. The table face which can be inclined and folded vertically is applicable to video watching of a single person or multiple people. The touch table comprises the table face, a touch screen displayer, a base and a central computer. The touch screen is embedded in the table face, and the table face is disposed above a bottom platform to be connected with the bottom platform. An audible device is mounted on the side wall of the base, the length of the touch table face is equal to the width, and the side length is larger than 65cm and smaller than 125cm. According to the touch table, a plurality of directional or small-visual-angle film or display units are used and are laid horizontally on the surface of a display panel, a plurality of palm terminals are networked with a touch table central computer through a wireless network to form a distributed game system. The touch table face is in direct or indirection connection with the bottom platform through a rotation shaft, the touch table face can be rotated to be disposed slantwise or vertically, and a supporting rod plays a role in supporting the table face.

The invention discloses a near field sound source real-time positioning system based on a microphone array, belonging to the technical field of sound source positioning. The positioning system comprises a microphone array, an amplifying circuit module, a signal acquiring and processing module, a positioning result storage module and other system assistance modules. Voice signals are amplified through the microphone array, and the amplifying circuit adopts a special microphone amplifier with an automatic gain function as a core. The signal acquiring and processing module takes a DSP chip with an on-chip analog-to-digital conversion function as a core, and adopts the direct storage access function in the DSP chip. A near-field two-dimensional broadband MUSIC algorithm is further optimized, the acceptation or rejection of the microphone array signal frequency spectrum sub-band is carried out, a sub-band matrix is subjected to MUSIC spectrum estimation, the computation burden of DSP chip is reduced, and the system instantaneity is improved. The sound source positioning result storage module adopts an SD card as a storage medium, the application range of the system is improved, and thecluster chain of speaker position data is constructed by adopting a 'pre-built cluster chain method', so that the system instantaneity is improved.

The invention discloses a method for ranking turbine rotor blades based on Monte Carlo tree search method for ranking turbine rotor blades based on Monte Carlo tree search, belonging to the mechanicalassembly field. Firstly, the leaf parameters are extracted. Then, the root node of Monte Carlo tree is initialized, and different leaf sequences are generated by selecting nodes, expanding nodes andsimulating nodes for many times, and the unbalance quantity is calculated, and the weight value in Monte Carlo tree is updated. Secondly, the node with the largest weight and the corresponding optimalsequence are taken as output, the root node of the Monte Carlo tree is updated again, a new Monte Carlo tree is created, and the above process is repeated. Finally, the optimal unbalance quantity andits corresponding blade sequence are obtained by searching multiple Monte Carlo trees, and the rotor blade scheduling is optimized. The invention can further optimize the blade scheduling problem while ensuring good convergence, and the algorithm has good stability.

A release method and system for the relevant information of eh users can be established by using the input platform, collecting and storing customer input content, analyzing the input content in certain valve scope, deciding the relevant information to be sent to the user based on the analytical result, displaying the said relevant information on the computer equipment that runs on the input method platform. It can greatly improve the accuracy of related information, with the terminal end being the user, with greatly expanded information release expansion, to realize the information release directly facing the user. It can also select parameter information input by the users, combing the said parameter to analyze the input content within certain scope, to more accurately acquire the analytical result of the input content.