Abstract: Disclosed in the present application are an action recognition method and apparatus. The method comprises: acquiring a video clip, and determining at least two target objects in the video clip; for each of the at least two target objects, connecting positions of the target object in various video frames of the video clip, so as to construct a spatiotemporal graph of the target object; dividing at least two spatiotemporal graphs, which are constructed for the at least two target objects, into a plurality of spatiotemporal graph subsets, and determining a finally selected subset from the plurality of spatiotemporal graph subsets; and determining an action category of the action between the target objects that is indicated by a relationship between the spatiotemporal graphs included in the finally selected subset as the action category of an action included in the video clip.

Abstract: The present disclosure relates to a broadcast directing method, apparatus and system, and a non-transitory computer-storable medium, which relate to the technical field of computers. The broadcast directing method includes acquiring a reference video stream from a reference camera position; performing event recognition on the reference video stream to obtain at least one reference event frame interval; determining a local sequence of each reference event frame interval according to a correspondence relationship between the event and a camera position identification, the local sequence comprising a camera position identification of each frame image of a video to be played corresponding to the reference event frame interval and a frame identification corresponding to the camera position identification; generating a broadcast directing sequence according to the local sequence; and generating a broadcast directing video according to the broadcast directing sequence and a video stream of a camera position.

Abstract: This application provides a method and an apparatus, for accelerating cold startup of an application. The method includes after identifying an event that instructs an operating system of a terminal to cold start up an application, obtaining, from a plurality of dimensions, current status information related to the cold startup of the application, where the current status information includes a hardware configuration of the terminal, current load of the operating system of the terminal, resource overheads for cold starting up the application, and duration corresponding to each of a plurality of tasks in a process of cold starting up the application. The method also includes determining, by analyzing the current status information, a plurality of objects that need to be optimized in the current process of cold starting up the application; and then obtaining, based on the determined objects.

Abstract: The present disclosure provides an information advertisement method, a controller and a storage medium. The information advertisement method includes: determining, by a central controller, an advertisement domain; and advertising, by the central controller, a node in the advertisement domain of computing information. According to the present disclosure, it is able to reduce the amount of futile notification information in a network, and achieve fine management on a notification range of network information, thereby to improve a network resource utilization rate and an active bandwidth utilization rate.

Abstract: The present disclosure relates to the technical field of image processing, and in particular to an image description generation method, apparatus and system, and a medium and an electronic device.

Abstract: Provided in the present disclosure are a display substrate, a display panel, and a display device. The display substrate includes: a base substrate; a plurality of gate lines located on a side of the base substrate, and extending in a first direction; and a plurality of pixel electrodes, distributed in arrays, where pixel electrodes in the same row are electrically connected to at least one gate line, and in the pixel electrodes in the same row, at least adjacent parts of two adjacent pixel electrodes are distributed in a split-level manner.

Abstract: Disclosed is a prefabricated laminated slab made in accordance with a pretensioning fibermesh anti-cracking process and a manufacturing method thereof, and belongs to the technical field of fabricated structure engineering. The prefabricated laminated slab comprises a concrete layer, fiber grids, distribution bars and truss bars. The fiber grids are subjected to prestress tensioning and pouring to form a novel laminated slab with the pretensioning process, an overall pouring process of layered formworks is adopted, the construction problem during prestress tensioning of the fiber grids of the component can be effectively solved, the crack resistance of the laminated slab can be improved by adopting the prestress grids, the rigidity is increased, and the crack width under normal use conditions is reduced.

Abstract: A display panel includes a substrate, a pixel circuit layer, a light-emitting element, and an auxiliary electrode. The pixel circuit layer is arranged on a side of the substrate and includes a pixel circuit. The pixel circuit includes an initialization transistor. An active layer of the initialization transistor includes a channel region, a first electrode region, and a second electrode region. The light-emitting element is arranged on a side of the pixel circuit layer away from the substrate and electrically connected to the pixel circuit. The light-emitting element includes a first electrode, a light-emitting material layer, and a second electrode laminated on each other. The auxiliary electrode is arranged on the side of the pixel circuit layer away from the substrate and electrically connected to the second electrode. The auxiliary electrode is electrically connected to the second electrode region through a via hole.

Abstract: The display panel comprises: a plurality of pixels arranged in an array. Each pixel comprises a plurality of subpixels. The plurality of pixels arranged in an array comprise: a plurality of pixel rows. Each pixel row comprises a plurality of subpixels arranged in a first direction. The plurality of pixel rows extend in a second direction. The first direction intersects with the second direction. The display panel comprises: a first array substrate and a first opposing substrate arranged opposite to each other, and a first liquid crystal layer located between the first array substrate and the first opposing substrate. The first array substrate comprises: a plurality of first driving transistors arranged in an array. At least one pixel is arranged between two adjacent first driving transistors in the first direction. The first opposing substrate comprises: a plurality of light shielding portions in one-to-one correspondence to the first driving transistors.

Abstract: The present invention relates to the field of genetic engineering, particularly to a recombinant expression vector for rapidly screening the high expression strains and a method for rapidly screening high expression strains. In the invention, an exogenous red fluorescent protein and Aspergillus fumigatus cell surface protein localization signal are fused and expressed, and the fusion gene (DsRed-AfMP1) is integrated into the genome of Trichoderma reesei, so as to construct a strain displaying red fuorescent protein on the surface of Trichoderma reesei. By sorting Trichoderma reesei strains with red fluorescent protein on the surface by flow cytometry, genes beneficial to the improvement of cellulase activity can be quickly isolated.

Abstract: A method for identifying an individual of an oplegnathus punctatus based on a convolutional neural network is provided. Target initial positioning involves three continuous convolutional layers and an average pooling layer. A region of feature interest point is obtained, a hyperparameter candidate box is set to obtain a region, thereby obtaining an approximate position of a target object. II_Net backbone convolutional neural network includes six convolutional layers and four pooling layers, which includes a LeakyReLU activation function used as an activation function of the first convolutional layer, convolutional network layers of Alexnet and parameter data, and a maximum pooling layer of an overlapped pooling structure. Fully connected layers use a genetic algorithm to improve data transmission between layers. With the established model, identification of the individual of the oplegnathus punctatus is performed by using test data.

Abstract: The invention provides a laser processing system having the function of optical diffraction tomography, comprising: an integrated imaging optical path and a processing optical path; the imaging optical path is used to perform optical diffraction tomography on the device to be processed; the processing optical path is used for processing the device to be processed. Moreover, the specific optical path structure is introduced. During the processing of the device to be processed, the laser processing system can also perform real-time imaging of the device to be processed without shifting the device to be processed. That is to say, the laser processing and imaging processing of the device to be processed can be realized at the same time in a laser processing system.

Abstract: By the abovementioned technical solutions, the present disclosure provides optimizing a vector autoregressive model conforming to structural constraints of sparsity and acyclicity. A regularization term is introduced to the model to impose the sparsity structural constraint such that most off-diagonal coefficients of an autoregressive coefficient matrix are forced to zero values. One or more penalty terms are introduced to the model to impose the acyclicity structural constraint such that coefficients of the main diagonal are not causally self-related. The resulting model is then reformulated for computation as an augmented Lagrangian function, and further computed for different parameters in alternating iterations to make the computations tractable and within magnitude and precision limits of digital computers.

Abstract: A method for identifying an individual of an oplegnathus punctatus based on a convolutional neural network is provided. Target initial positioning involves three continuous convolutional layers and an average pooling layer. A region of feature interest point is obtained, a hyperparameter candidate box is set to obtain a region, thereby obtaining an approximate position of a target object. II_Net backbone convolutional neural network includes six convolutional layers and four pooling layers, which includes a LeakyReLU activation function used as an activation function of the first convolutional layer, convolutional network layers of Alexnet and parameter data, and a maximum pooling layer of an overlapped pooling structure. Fully connected layers use a genetic algorithm to improve data transmission between layers. With the established model, identification of the individual of the oplegnathus punctatus is performed by using test data.

Abstract: Time series decomposition method and system are provided. Time series data including at least a seasonality component is received. A noise removal filter may be applied to the time series data. Furthermore, a trend component and the seasonality component are extracted from the time series data. A residual component may subsequently be extracted from the time series data based on the trend component and the seasonality component. Anomaly detection may then be performed on the trend component or the residual component to determine whether an anomaly exists in the time series data.

Abstract: Input time series data of a particular time length may be received, and pre-processed to obtain processed time series data. The processed time series data may then be decomposed into a plurality of pieces of time series data of the particular time length at different levels, and a detection of whether a piece of time series data at a particular level includes a periodic component candidate in a frequency domain may be performed. In response to detecting that the piece of time series data at the particular level includes the periodic component candidate in the frequency domain, a validation may be performed to determine whether the periodic component candidate in the piece of time series data at the particular level is a true periodic component of the input time series data in a time domain.

Abstract: By the abovementioned technical solutions, the present disclosure provides optimizing a vector autoregressive model conforming to structural constraints of sparsity and acyclicity. A regularization term is introduced to the model to impose the sparsity structural constraint such that most off-diagonal coefficients of an autoregressive coefficient matrix are forced to zero values. One or more penalty terms are introduced to the model to impose the acyclicity structural constraint such that coefficients of the main diagonal are not causally self-related. The resulting model is then reformulated for computation as an augmented Lagrangian function, and further computed for different parameters in alternating iterations to make the computations tractable and within magnitude and precision limits of digital computers.

Abstract: Input time series data of a particular time length may be received, and pre-processed to obtain processed time series data. The processed time series data may then be decomposed into a plurality of pieces of time series data of the particular time length at different levels, and a detection of whether a piece of time series data at a particular level includes a periodic component candidate in a frequency domain may be performed. In response to detecting that the piece of time series data at the particular level includes the periodic component candidate in the frequency domain, a validation may be performed to determine whether the periodic component candidate in the piece of time series data at the particular level is a true periodic component of the input time series data in a time domain.

Abstract: Time series decomposition method and system are provided. Time series data including at least a seasonality component is received. A noise removal filter may be applied to the time series data. Furthermore, a trend component and the seasonality component are extracted from the time series data. A residual component may subsequently be extracted from the time series data based on the trend component and the seasonality component. Anomaly detection may then be performed on the trend component or the residual component to determine whether an anomaly exists in the time series data.

Abstract: An electrical connector includes: an insulative housing comprising a base portion defining a rearward surface and a tongue portion extending forwardly from the base portion; plural conductive terminals affixed to the insulative housing and each having a contacting portion exposed to the top and bottom surfaces of the tongue portion, a fixing portion embedded in the base portion, and a soldering portion extending backwardly out of the base portion; a shielding plate affixed to the insulative housing; a shielding shell covering the insulative housing and defining a sol space with the rearward surface of the base portion; and a waterproof sheet; wherein the base portion further comprises a stepped portion in the sol space, the waterproof sheet is formed by solidification of liquid insulative material flowing from the stepped portion to the rearward surface, and the waterproof sheet encloses the stepped portion.