Abstract: Disclosed is a dynamic network risk predicting method based on a graph neural network, which comprises the following steps: collecting network data firstly and constructing a network data sequence graph; extracting the time sequence of the network data sequence graph and obtaining a time sequence feature graph; extracting network attribute features, network structure features and network change features of the dynamic network; obtaining the representation vector of the network sequence graph by learning the representation vector; constructing the anomaly detection model and carrying out the dynamic network risk prediction.

Abstract: Provided are a speech translation method, a device, and a storage medium. The method includes: extracting, through an encoder of an end-to-end speech translation model, the semantic feature of a to-be-processed speech; decoding, through a decoder of the end-to-end speech translation model, a source language text corresponding to the semantic feature from the semantic feature; decoding, through the decoder of the end-to-end speech translation model, the semantic feature according to the source language text to obtain a text sequence corresponding to the semantic feature; and splitting the text sequence to obtain a target language text corresponding to the to-be-processed speech.

Abstract: A battery cell, including an electrode assembly and a housing accommodating the electrode assembly. The housing includes a cavity and a cover covering the cavity. The cavity includes a flange extending along a border of an opening of the cavity. Periphery of the cover is connected to the flange to form a side wing to seal the cavity. The battery cell further includes a tab sandwiched between the side wings and extending out of the housing along the side wings. The side wing has a first sealing structure and a second sealing structure. A direction from an edge, away from the border, of the side wing to the border is defined as a first direction, and the second sealing structure exceeds the first sealing structure in the first direction.

Abstract: An array substrate, a display panel and a display device are provided. The array substrate includes a plurality of pixel circuits arranged in an array in a first direction and a second direction, where each pixel circuit includes a first semiconductor portion including a first node and a second node; a first reference signal line and a second reference signal line electrically connected to the first node and the second node, respectively; and a switch element formed by a second semiconductor portion and a control portion. In adjacent two pixel circuits in the second direction, the first node of one pixel circuit is connected to the second node of the other one pixel circuit through the second semiconductor portion. The control portion has an orthographic projection on a plane of the array substrate at least partially overlapping the second semiconductor portion.

Abstract: Provided are a method and apparatus for determining a security area, a device, and a storage medium. The method includes: determining property information of an entity in a real scene in which a user is in a motion; displaying, based on the property information of the entity, in a virtual space a target virtual scene mapped from the real scene; and determining, based on a security-type virtual object in the target virtual scene, a corresponding security area. In the present disclosure, the customized delineation of the security area in a plurality of spatial ranges is achieved by using the user's motion in the real scene, and an adaptability between the security area and the real scene is ensured. Moreover, by displaying the target virtual scene mapped from the real scene to the user, the user's collision-free motion in the real scene can be supported before the security area is delineated.

Abstract: A traffic forwarding method and an apparatus, related to the field of communication technologies, to reduce a rate of traffic loss. The method includes: during graceful restart of a device in a network, when a restart device in the network enters a restart state, a neighbor device of the restart device updates, from first routing information to second routing information, routing information used to forward target traffic. In addition, the neighbor device forwards the target traffic based on the second routing information. The first routing information indicates the neighbor device to send the target traffic to the restart device. The second routing information indicates the neighbor device to send the target traffic to another device.

Abstract: The present disclosure relates to a media retrieval method and apparatus, wherein the method comprising: acquiring a media feature of a query media as a first media feature which contains a plurality of first media feature units; performing, according to each individual of the first media feature unit, a first ranking on a plurality of existing media, and taking, according to a result of the first ranking, first k existing media as a first candidate media set, where k is a positive integer; and performing, according to the plurality of the first media feature units arranged in sequence, a second ranking on the first candidate media set, and taking, according to a result of the second ranking, first n first candidate media as a retrieval result, where n is a positive integer.

Abstract: The present disclosure provides an image processing method and apparatus, a device, and a medium. The method is applied in a virtual reality headset and includes: obtaining six-degree-of-freedom data of the virtual reality headset and a virtual reality image corresponding to the six-degree-of-freedom data; determining a motion vector of a user in a target direction based on the six-degree-of-freedom data; determining a to-be-processed object from the virtual reality image; and scaling the to-be-processed object based on the motion vector. In the present disclosure, stereoscopic scaling of a VR image displayed by a VR device is implemented to achieve a stereoscopic scaling effect of the image, thereby satisfying use requirements for a user.

Abstract: The present invention relates to the field of inland vessel identification and ranging technology, and discloses a method, system, medium, equipment and terminal for identifying and ranging inland vessels. In the stage of vessel identification, based on the classical YOLO-V4 network model, the MobileNetV1 network is used to replace the feature extraction network CSPDarknet53 of the YOLO-V4 model; In the stage of vessel ranging, a binocular stereo vision ranging model is established, and the FSRCNN is used for super-resolution reconstruction of the original image pairs to enhance the vessel feature information; the ORB algorithm is used to achieve feature detection and matching at the sub-pixel level to obtain the parallax value between image pairs, and the depth information of the vessel target is obtained by triangulation principle and coordinate conversion.

Abstract: The permanent magnet comprises a main phase structure of R2T14B crystal grains, and R is a rare earth element; T comprises at least Mn, Fe, and optionally a transition metal comprising Co; B is boron; the permanent magnet further comprises Mn and heavy rare earth elements which are distributed in a grain boundary in a diffusion mode. The heavy rare earth element is selected from at least one selected from Dy, Ho and Tb. According to the rare earth permanent magnet prepared through the preparation method, more heavy rare earth elements can be diffused into the magnet core along the grain boundary, Hcj distribution of the permanent magnet is improved, and meanwhile the corrosion resistance and the mechanical property of the permanent magnet are improved.

Abstract: Programmable pixel blending method and apparatus, computer device, storage medium and computer program product are provided. The method includes: performing rasterization on a first triangle, a second triangle, and a third triangle via the raster unit to obtain triangle coverage information of each triangle; sending the triangle coverage information to the warp assembly unit and the warp reorder unit respectively; assembling, by the warp assembly unit, the first triangle and the second triangle into a first warp, and requesting a warp from a warp request interface based on information of the first warp; executing the first warp based on a warp information cache, a warp synchronization unit, the warp reorder unit, the pixel operation unit, the interpolation unit, and the execution unit; after the first warp is executed, continuing to assemble, by the warp assembly unit, the third triangle and completing a rendering of the third triangle.

Abstract: A thin film transistor and a manufacturing method therefor, an array substrate, and a display device. The thin film transistor includes an active layer, a gate insulating layer, and a gate electrode; the gate insulating layer is located on one side of the active layer; the gate electrode is located on one side of the gate insulating layer distant from the active layer; the gate electrode includes an opening a part of the active layer overlapped with the opening includes a first lightly doped region, a first heavily doped region, and a second lightly doped region that are sequentially arranged along a first direction parallel to a plane where the active layer is located.

Abstract: A casting and outputting system is provided. The system includes a casting machine, a disc body, a transfer device, a special fixture, a cooling tank, a collection device and a waste plate rack. The transfer device is configured to drive the special fixture to take out anode plates cast on the disc body, place a qualified one of the anode plates in the cooling tank, place an unqualified one of the anode plates on the waste plate rack, lift the anode plate cooled in the cooling tank, and place the cooled anode plate on the collection device for stacking.

Abstract: Exemplary embodiments include a system having a first wireless audio output device configured to connect to a source device via a first piconet and a second wireless audio output device configured to connect to the first wireless audio output device via a second piconet. A schedule of the first piconet includes a plurality of slots associated with an audio packet, a first subset of the slots used by the source device to transmit the audio packet, the first and second wireless audio output devices tuning to the first piconet to listen for the transmissions of the audio packet, and when, after a last one of the first subset of slots, the first or second wireless audio output devices did not receive the audio packet, the first and second wireless audio output devices exchange information via the second piconet such that the both wireless audio output device receive the audio packet.

Abstract: A power supply circuit that may be operated in a pre-charge mode or a switch-charge mode to provide a charging current. In the pre-charge mode, a current sense transistor of the power supply circuit may be operating to sense the charging current and to provide a current sense signal indicative of the charging current, while a current control transistor of the power supply circuit may be operating to control the charging current based on the current sense signal.

Abstract: The present disclosure provides an embedded mounting bracket and camera assembly comprising same. The embedded mounting bracket is used for mounting a camera, and the embedded mounting bracket includes a bracket body and a mounting member. The bracket body includes a bracket side wall, the bracket side wall includes a lower end surface and an upper end surface opposite to the lower end surface, and the upper end surface includes multiple step surfaces with different heights from the lower end surface. The mounting member is detachably assembled to the bracket body for bearing the camera. The mounting member is selectively assembled on the step surfaces with different heights, so as to replaceably assemble at least two cameras with different body heights to the bracket body.

Abstract: A structure includes a heat dissipation base having through bores formed at four corners thereof for screw fastening elements to extend therethrough and fix the heat dissipation base to a top of a heat source, a boss upward protruded from the heat dissipation base, and a shaft extended in a radial direction through the boss with two ends projected from the boss; a hold-down member disposed on the heat dissipation base and formed with a through-opening for externally fitted around the boss; and an adjustment element having a hollow cylindrical body fitting around the boss and having a sliding slot section and a force applying section operable for the sliding slot section and the shaft to slide relative to each other and drive the hollow cylindrical body downward against the hold-down member, forcing the latter to apply evenly downward forces on a central area of the heat dissipation base.

Abstract: This invention proposes a single-vector-based finite control set model predictive control for two parallel power converters, which adopts a centralized control structure to achieve accurate control of overall performance. It establishes predictive models for line currents and three phase-circulating currents and constructs a novel cost function that uses these currents as performance indices to implement the predictive control algorithm based on the proposed predictive models. The invention proposes dynamic weighting coefficients and adjustment principles to improve system control performance. A finite set output signal matrix containing important characteristic information of all alternative vectors is constructed to avoid redundant calculations in each control horizon, reducing computation time during practical implementation.

Abstract: Provided are a point cloud segmentation method and apparatus, a device, and a storage medium. The point cloud segmentation method includes: acquiring a to-be-processed point cloud; obtaining, in a gridding scenario space to which respective point in the point cloud belongs, a target grid corresponding to the respective point through a pre-trained neural network, wherein the pre-trained neural network is obtained by training a sample point cloud and a sample target grid corresponding to the sample point cloud in a sample gridding scenario space; and outputting a point cloud corresponding to a respective instance according to an instance category corresponding to the target grid, wherein the same target grid has the same instance category.

Abstract: Methods performed by a first sink device, a source device, or a second sink device. The first sink device is connected to a source device via a first communication link and a second sink device via a second communication link, wherein the second sink device is configured to eavesdrop on communications between the first sink device and the source device on the first communication link. The methods include determining an occurrence of a trigger event and modifying an operation of at least one of the first sink device, the second sink device or the source device based at least on the trigger event occurring.