Abstract: An object detection training method can include receiving a training sample set in a current iteration of an object detection training process over an object detection neural network. The training sample set can include first samples of a first class and second samples of a second class. A first center loss value of each of the first and second samples can be determined. The first center loss value can be a distance between a feature vector of the respective sample and a center feature vector of the first or second class which the respective sample belongs to. A second center loss value of the training sample set can be determined according to the first center loss values of the first and second samples. A first target loss value of the current iteration can be determined according to the second center loss value of the training sample set.

Abstract: A role simulation method and a terminal apparatus in a VR scene are provided. The role simulation method in the VR scene includes: obtaining a role obtaining instruction triggered by a hand model in the VR scene, movement of the hand model in the VR scene being controlled by an interaction controller; determining a virtual role from the VR scene according to the role obtaining instruction; dynamically adjusting virtual props in the VR scene according to a music file played in the VR scene; displaying the virtual role in the VR scene through a VR display, and displaying the virtual props that are dynamically adjusted.

Abstract: An event execution method includes executing a first event by a first application client logged in on a mobile terminal of a user by using a first account of the user to obtain first event result information, the first event result information indicating an account resource to be deducted, displaying the first event result information on a screen of a processing device, and requesting to execute a second event by using the processing device, the second event being to deduct an amount of resource from a second account associated with the first account.

Abstract: An image processing method and apparatus, and a computer readable medium are provided. The method includes obtaining an image. The image is processed using a preset training model that is a function relationship model of a feature sample image and an activation function of the feature sample image. The feature sample image includes an image satisfying an image feature value extraction condition. A target image is obtained that corresponds to the image according to a processing result of the preset training model.

Abstract: A resource-push method is disclosed. The method includes obtaining, by a resource-push server, a target relationship chain of a benchmark user; determining at least one relational user of the benchmark user according to the benchmark user and the target relationship chain; and obtaining parameter characteristics of the at least one relational user from a preset database. The method also includes determining a similarity value between each of the at least one relational user and the benchmark user according to the parameter characteristics, or a probability value of each of the at least one relational user according to the parameter characteristics; and determining at least one push user from the at least one relational user according to the similarity value or the probability value, so as to push a target resource to the at least one push user.

Abstract: A client connection method is provided. The method includes obtaining, by a second client, a first link from a first sharing identifier of a first client. The first link includes a first identification code and service type information, the first identification code is used for identifying the first client, and the service type information represents a service type corresponding to a client connection server. The second client extracts the first identification code and the service type information according to the first link, determines the client connection server according to the service type information, and sends a connection request carrying the first identification code and a second identification code to the client connection server. The second client connects with the first client based on a binding connection between the first client and the second client established by the client connection server.

Abstract: The present disclosure discloses a method and an apparatus for processing a group chat message. The method includes: determining that a specified user is joined into a group; if at least one chat message exists in the group within a preset time period before a current time, obtaining the at least one chat message within the preset time period; selecting a chat message from the at least one chat message within the preset time period; and sending a message identifier of the selected chat message to a target server, so that the target server sends the selected chat message in the group according to the message identifier of the selected chat message. According to the present disclosure, by means of a simple selection operation, a previous chat message can be presented to a specified user that newly joins a group, thereby greatly improving processing efficiency.

Abstract: A video classification method is provided for a computer device. The method includes obtaining a to-be-processed video, where the to-be-processed video has a plurality of video frames, and each video frame corresponds to one time feature; and sampling the to-be-processed video according to a time-feature sampling rule, and obtaining at least one video frame feature sequence. The time-feature sampling rule is a correspondence between time features and video frame feature sequences. The method also includes processing the video frame feature sequence by using a first neural network model, to obtain a feature representation result corresponding to the video frame feature sequence, where the first neural network model is a recurrent neural network model; and processing the feature representation result by using a second neural network model, to obtain a prediction result corresponding to the video frame feature sequence.

Abstract: This application is related to a data recovery system. The data recovery system includes processing circuitry which receives a restore request. The restore request includes a restore target time. When backup data corresponding to the restore target time is determined to be available, the processing circuitry creates a temporary database service instance and determines information of the temporary database service instance. The processing circuitry further obtains the backup data corresponding to the restore target time and writes the backup data into the temporary database service instance according to the information of the temporary database service instance. The processing circuitry executes a formal restoration operation on the temporary database service instance, in which the backup data is written, according to a received processing request.

Abstract: A cloud-end data multicast method includes: obtaining a multicast packet, the multicast packet carrying a tenant identifier, a destination address, and a source address; and searching for a multicast group according to the tenant identifier and the destination address. The multicast group includes multicast members. The method also includes obtaining routes corresponding to the multicast members; generating a routing tree according to the routes corresponding to multicast members; and obtaining member addresses corresponding to the multicast members. The method also includes obtaining an address list containing addresses of target members that are identified to receive the multicast packet by performing address filtering according to the source address and the member addresses; encapsulating the multicast packet; and delivering the encapsulated multicast packet to the target members according to the address list and the routing tree.

Abstract: The present disclosure discloses a method and an apparatus for distinguishing objects. The method includes: obtaining a plurality of object groups displayed in an image, each object group including at least one target object, and a same resource being configured for target objects in different object groups; setting different mark values for the plurality of object groups, target objects in a same object group having a same mark value; and separately performing pixel correction on pixels of the target objects in each object group according to the mark value of each object group, pixels of the target objects having different mark values being corrected to have different display attributes.

Abstract: This application discloses a data processing method, system, and apparatus, a storage medium, and a device, and belongs to the field of database technologies. The method includes receiving, a trigger request; triggering, according to the trigger request, the first cloud encryptor to store a root key seed, an operating policy, a data key seed, and a data key identifier, and triggering the database proxy to store an encryption data dictionary, the operating policy indicating an operation policy of the first cloud encryptor. The method further includes receiving a data processing request from the client; sending first data that the data processing request requests to process and the data key identifier in the encryption data dictionary to the first cloud encryptor. The method further includes implementing the operating policy, processing the first data, and responding to the data processing request by using the second data.

Abstract: A method is provided for displaying a page. The method includes extracting, from a to-be-displayed page, page parameter information of the to-be-displayed page. The page parameter information includes at least one of a uniform resource locator (URL) of the to-be-displayed page, a type of page content, and a page domain name. The method also includes determining, according to the page parameter information of the to-be-displayed page, a display mode to display the to-be-displayed page; and displaying the to-be-displayed page according to the determined display mode.

Abstract: The present disclosure provides a method for identifying a clique network using a Pregel graph computing framework. The method includes determining, according to a data transfer feature of a target clique, a transmitting direction of attribute values of nodes; transmitting, an edge vector of the edge and an attribute value of the first node to the second node along the transmitting direction in a constructed data transmission relationship network, the edge vector comprising a plurality of data transfer eigenvalues; performing, weighted calculation on the edge vector received by the second node to obtain an optimal weighted edge; iterating, the above operations; and determining, according to attribute values of the nodes after the one or more iterations, nodes in the target clique, and determining attributes of the nodes in the target clique.

Abstract: This application discloses a flight tag obtaining method, terminal, and server. A takeoff geographic location of a first unmanned aerial vehicle can be obtained when the first unmanned aerial vehicle takes off. The takeoff geographic location of the first unmanned aerial vehicle can then be sent to a server. A flight tag returned by the server can be obtained. The flight tag can indicate whether another aerial vehicle took off from the takeoff geographic location of the first unmanned aerial vehicle before the first unmanned aerial vehicle.

Abstract: Provided are a gesture control method, a gesture control device and a gesture control system. The gesture control method includes: establishing, by a mobile terminal, a wireless communication connection with a display device; displaying, by the mobile terminal, a gesture input interface corresponding to a current application of the display device; detecting, by the mobile terminal, a user gesture; converting, by the mobile terminal, the user gesture into an operation instruction corresponding to the user gesture; and transmitting, by the mobile terminal, the operation instruction to the display device.

Abstract: An image recognition method, a terminal, and a storage medium are provided. The method includes: performing feature extraction on a to-be-recognized image by using an encoder, to obtain a feature vector and a first annotation vector set; performing initialization processing on the feature vector, to obtain first initial input data; and generating first guiding information based on the first annotation vector set by using a first guiding network model. The first guiding network model is configured to generate guiding information according to an annotation vector set of any image. The method also includes determining a descriptive statement of the image based on the first guiding information, the first annotation vector set, and the first initial input data by using a decoder.

Abstract: Natural language question answering method and apparatus in the field of information retrieval and processing are provided. The method includes: acquiring a natural language question N; converting the natural language question N into a query semantic graph Qs, each edge in the query semantic graph Qs representing one semantic relation in the natural language question N; searching an RDF graph G for a subgraph matching the query semantic graph Qs; and obtaining an answer to the natural language question N according to the subgraph.

Abstract: Options are displayed on a user interface that provides a network service, a selection instruction of a user is received, and an option selected by the user is acquired according to the selection instruction; the option selected by the user is recorded; and an advertisement that matches the option is queried for, and the advertisement is presented to the user.

Abstract: The present disclosure discloses a method for plugin loading. The method includes obtaining an identifier of a plugin component of a plugin from a threading module, recording the identifier of the plugin component, and replacing the identifier of the plugin component with an identifier of a host component of an application program. The method also includes sending the identifier of the host component to a threading module, to perform system permission verification and receiving runnable notification information when passing the verification. The method further includes in response to the received runnable notification information, replacing the identifier of the host component with the identifier of the plugin component according to the recorded identifier of the plugin component; and sending the identifier of the plugin component to the threading module, to load the plugin. The present invention further discloses an apparatus and a storage medium for plugin loading.