Abstract: An image processing method includes: obtaining a target image; performing feature extraction on the target image based on a residual network, to obtain image feature information; and performing recognition processing on the target image according to the image feature information. The residual network includes a plurality of residual blocks that are successively connected, each of the residual blocks including a convolution branch and a residual branch, a size of a convolution kernel of a first convolutional layer in the convolution branch being less than a size of a convolution kernel of a second convolutional layer located after the first convolutional layer, and a convolution stride of the second convolutional layer being greater than a convolution stride of the first convolutional layer and less than a width of the convolution kernel of the second convolutional layer.

Abstract: A processing method that is performed by one or more processor is provided. The processing method includes determining a target video frame in a currently captured video; determining an object area in the target video frame based on a box selection model; determining a category of a target object in the object area based on a classification model used to classify an object in the object area; obtaining augmented reality scene information associated with the category of the target object; and performing augmented reality processing on the object area in the target video frame and the augmented reality scene information, to obtain the augmented reality scene.

Abstract: A depth image generation method, apparatus, and storage medium and electronic device. The method includes: acquiring a plurality of target images; performing multi-stage convolution processing on the plurality of target images through a plurality of convolutional layers in a convolution model to obtain feature map sets respectively outputted by the plurality of convolutional layers; performing view aggregation on a plurality of feature maps in each feature map set respectively to obtain an aggregated feature corresponding to each feature map set; and performing fusion processing on the plurality of obtained aggregated features to obtain a depth image. The plurality of acquired target images are obtained by photographing the target object from different views respectively, so that the plurality of obtained target images include information from different angles, which enriches information content of the acquired target images.

Abstract: A method for establishing a voice activity detection model includes obtaining a training audio file and a target result of the training audio file, framing the training audio file to obtain an audio frame, extracting an audio feature of the audio frame, the audio feature comprising at least two types of features, inputting the extracted audio feature as an input to a deep neural network model, performing information processing on the audio feature through a hidden layer of the deep neural network model, and outputting the processed audio feature through an output layer of the deep neural network model, to obtain a training result; determining a bias between the training result and the target result, and inputting the bias as an input to an error back propagation mechanism, and updating weights of the hidden layer until the deep neural network model reaches a preset condition.

Abstract: Provided is an anonymous account authentication method, apparatus and storage medium. The method may include receiving, from a user, an anonymous account authentication request, requesting a resource owner to perform authentication on a use permission of a resource using an anonymous account, obtaining a first version number of the resource owner according to the anonymous account authentication request, the first version number representing an identification number of a current version of an authentication policy table, obtaining a linear policy table based on the first version number matching a second version number, the second version number representing an identification number of a current version of the linear policy table, obtaining an authentication policy set associated with the resource from the linear policy table, and performing authentication by using the authentication policy set, the authentication policy set comprising at least one authentication policy.

Abstract: A virtual character control method is provided. The method includes displaying a user interface (UI) of an application program, the UI including a virtual environment image and a movement control superimposed on the virtual environment image, and the virtual environment image being an image of a virtual environment, including a virtual character, observed with a perspective; receiving a pressure operation with respect to the movement control, the pressure operation having a pressure greater than a pressure threshold; and controlling, based on the pressure operation satisfying a trigger condition, the virtual character to perform a target action in the virtual environment. Apparatus, terminal, and non-transitory computer-readable storage medium counterparts are also contemplated.

Abstract: Embodiments of the present disclosure relate to a content processing method and apparatus, a computer-readable storage medium, and a computer device. The method includes obtaining data of a to-be-pushed user in behavior evaluation dimensions, user portrait information of the to-be-pushed user, and to-be-pushed content, with the user portrait information corresponding to a user portrait code and the to-be-pushed content corresponding to a to-be-pushed content code. The method further includes determining a user behavior code based on the data of the to-be-pushed user in the behavior evaluation dimensions, and obtaining a target code by fusing the to-be-pushed content code with the user behavior code and the user portrait code. Then, the method includes determining a recommendation probability of the to-be-pushed content based on the target code to determine the push content based on the recommendation probability.

Abstract: A block chain-based data query method, performed by a server of a block chain system, is provided. In the method, a data identifier of first data to be queried is obtained according to a query request, the query request including a user identifier of a user that initiates the query request. Data permission information in a target block in a target block chain of the block chain system is obtained according to the data identifier. Based on the data permission information not including the user identifier, a query permission request is transmitted to at least one target server in the block chain system, the query permission request including the user identifier and the data identifier. Based on a received query permission response satisfying a target policy in the data permission information, the first data is obtained from a data storage system according to the data identifier.

Abstract: A method, an apparatus, a device and a system for session reestablishment and a computer storage medium. The method includes: transmitting a session reestablishment request of reestablishing a predetermined session to a session receiver, in a case that a session initiator determines that a predetermined session established with the session receiver is interrupted and the session receiver supports reestablishment of the predetermined session; and reestablishing, by the session initiator, the predetermined session with the session receiver according to a session reestablishment parameter of the predetermined session in a case that an acknowledgment for the session reestablishment request is received from the session receiver. The session reestablishment parameter is determined based on a status parameter of the predetermined session during interruption. Additionally, a method, an apparatus, a device and a system for session sharing, and a computer storage medium.

Abstract: This application belongs to the field of computer technologies, and provides an animation making method and apparatus, a computing device and a storage medium, to improve execution efficiency of animation making. In response to a pose selection instruction for a non-reference skeleton pose, a target plug-in node is invoked, the target plug-in node obtaining a non-reference skeleton shape model corresponding to the non-reference skeleton pose from a non-reference skeleton shape model set based on the pose selection instruction; target skeleton pose for an animated character is determined based on a parameter input instruction for a parameter of the target skeleton pose; and a target skeleton shape model of the target skeleton pose is generated based on the obtained non-reference skeleton shape model of the non-reference skeleton pose.

Abstract: A road updating method for an electronic map is provided. The method includes acquiring a target road in the electronic map, the target road being at least one of an intrinsic upper road of an overpass or an intrinsic lower road of the overpass, acquiring a target node of the target road, the target node being at least one of a first node of the target road or a last node of the target road, acquiring visible points within a preset range of the target node, determining path lengths between the visible points and the target node, determining an extended road of the target node according to the path lengths between the visible points and the target node, and updating the electronic map based on the extended road.

Abstract: An image processing method and device, storage medium and electronic device for deblurring an image. The method includes obtaining an image processing instruction including an instruction to deblur a target blurred image; obtaining a target model by training an original model based on a plurality of sample images of different scales, one of the plurality of sample images being a blurred image composed of a plurality of clear images, and the obtained target model being used for deblurring the blurred image to obtain a clear image; based on the image processing instruction, using the target model to deblur the target blurred image to obtain a target clear image; and outputting the target clear image.

Abstract: A video play method, a terminal device, and a storage medium for switching bitstream in a video play. The video play method includes obtaining a first image frame sequence and a second image frame sequence in response to a switch instruction to switch from a first video stream to a second video stream, the first image frame sequence in a form of the first video stream and the second image frame sequence in a form of the second video stream; determining a first image frame as a switching point in the first image frame sequence and a second image frame matching the first image frame in the second image frame sequence; determining the second image frame as a starting point in the second image frame sequence; and playing the first image frame sequence until the switching point, and playing the second image frame sequence from the starting point.

Abstract: Embodiments of the present disclosure include a method and apparatus for using a virtual prop, a device, and a storage medium, relating to supporting virtual environments. The method includes: displaying a user interface rom a first visual character's perspective; in response to a use instruction, controlling the first visual character to use a first virtual prop to generate a defense range an attack from another virtual prop; and intercepting the second virtual prop based on determining that the another virtual prop will likely fall within the defense range of the first virtual prop.

Abstract: Provided is a video processing method, including: obtaining a to-be-processed video and generating a first gait energy diagram, the to-be-processed video including an object with a to-be-recognized identity; obtaining a second gait energy diagram, the second gait energy diagram being generated based on a video including an object with a known identity; inputting the first gait energy diagram and the second gait energy diagram into a deep neural network; extracting respective identity information of the first gait energy diagram and the second gait energy diagram, and determining a fused gait feature vector from gait feature vectors of the first gait energy diagram and the second gait energy diagram; and calculating a similarity based on at least the fused gait feature vector. The identity information of the first gait energy diagram includes gait feature vectors, and the identity information of the second gait energy diagram includes gait feature vectors.

Abstract: Provided is a border detection method, server, and storage medium. The method including detecting a plurality of first straight line segments in a to-be-detected image, the to-be-detected image comprising a target region of a to-be-determined border; generating a plurality of first candidate borders of the target region according to the plurality of first straight line segments; obtaining a plurality of second candidate borders of the target region from the plurality of first candidate borders; extracting border features of the plurality of second candidate borders; and obtaining an actual border of the target region from the plurality of second candidate borders according to the border features of the plurality of second candidate borders and a border detection model, the border detection model being used to determine a detected value of each candidate border, and the detected value representing a similarity between each candidate border and the actual border.

Abstract: A method of processing data in a virtual reality (VR) application is provided. The VR application is restored to an initial running state in response to receiving a playback start instruction. Manipulation data that are transmitted to a runtime library to be pre-captured and stored are obtained, the manipulation data having a time sequence and being used to control running of the VR application. The manipulation data are imported into the runtime library based on the time sequence. The imported manipulation data are transferred to the VR application in the initial running state by using the runtime library and performing playback processing based on the transferred manipulation data.

Abstract: A road condition status prediction method is performed by a server and includes obtaining historical road segment information, the historical road segment information including road condition statuses and classification events that are used to classify the road condition statuses, and the classification events being determined based on content of the classification events, and obtaining an occurrence probability of each of the classification events, based on the obtained historical road segment information. The road condition status prediction method further includes obtaining a conditional probability of each of the classification events in each of the road condition statuses, based on the obtained historical road segment information, and predicting a road condition status of a road segment, based on the obtained occurrence probability of each of the classification events and the obtained conditional probability of each of the classification events in each of the road condition statuses.

Abstract: A method of controlling playing of multimedia is provided. The method includes obtaining, by a terminal, a first resource list of a first user interface (UI), the first resource list including one or more pieces of multimedia; obtaining, by the terminal based on detection of a trigger operation on first multimedia in the first UI, a resource type of the first multimedia; determining, by the terminal, a second resource list of a second UI based on the resource type of the first multimedia, the second resource list including at least the first multimedia; and switching, by the terminal, the first UI to the second UI based on the second resource list, and playing resource content of the first multimedia in the second UI.

Abstract: A method that is performed by a video encoding device is provided. The method includes identifying N image frames in a sliding window of an image frame sequence, the N image frames including N?1 encoded image frames and a to-be-encoded image frame; obtaining a motion amplitude difference of the N image frames in the sliding window, the motion amplitude difference indicating a difference between a corresponding image frame motion amplitude and a previous image frame motion amplitude; updating a static variable based on the motion amplitude difference of each of the N image frames in the sliding window, the static variable indicating a number of consecutive static image frames; and encoding the to-be-encoded image frame as an I frame based on the updated static variable not being less than a first preset threshold.