Abstract: A technique for reconstructing a three-dimensional scene from monocular video adaptively allocates an explicit sparse-dense voxel grid with dense voxel blocks around surfaces in the scene and sparse voxel blocks further from the surfaces. In contrast to conventional systems, the two-level voxel grid can be efficiently queried and sampled. In an embodiment, the scene surface geometry is represented as a signed distance field (SDF). Representation of the scene surface geometry can be extended to multi-modal data such as semantic labels and color. Because properties stored in the sparse-dense voxel grid structure are differentiable, the scene surface geometry can be optimized via differentiable volume rendering.

Abstract: The present disclosure belongs to the technical field of biological feeds, and in particular relates to use of an Areca catechu product as a sole active ingredient in manufacturing a biological feed and a dinking preparation for preventing and controlling African swine fever. In the present disclosure, the Areca catechu product as the sole active ingredient is mixed with an African swine fever virus (ASFV) suspension, and then inoculated in cells. Evaluation by in vitro cell experiments demonstrates that the Areca catechu product as the sole active ingredient has a good killing effect on ASFV, where 100% of ASFV with a virus titer of 5×103 HAD50/mL and can be killed within 30 minutes, and 100% of ASFV with a virus titer of 5×104 HAD50/mL can be killed within 2 hours.

Abstract: Various implementations disclosed herein include devices, systems, and methods that produce three-dimensional (3D) Gaussian data representing 3D appearance characteristics of surfaces depicted in 2D image content. For example, a process may obtain a 2D image representing a scene and determine depth information identifying depths of one or more portions of the scene represented in the 2D image. The process may further generate a representation comprising (3D) Gaussian data representing 3D appearance characteristics of the scene represented in the 2D image. The representation may be generated based on the 2D image and the depth information. The process may further provide the representation comprising the 3D Gaussian data for rendering multiple views of the scene from different viewpoints. The rendering may include generating splats based on the 3D Gaussian data for each of the different viewpoints.

Abstract: According to embodiments of the disclosure, a method, an apparatus, a device, and a storage medium for item recommendation are provided. A method includes: obtaining activity information of a target user within a target time period, the activity information being related to at least one type of interaction activity performed by the target user within the target time period, and a participation degree of the target user in the at least one type of interaction activity satisfying a preset condition; recognizing one or more target items associated with the target user by using a machine learning model and based on the activity information and sample information of the at least one type of interaction activity; and presenting recommendation information about the one or more target items.

Abstract: This application discloses a data update method and apparatus. The method includes: obtaining first data obtained by compressing third data in at least a first compression manner and a second compression manner successively, and second data obtained by compressing fourth data in at least the first compression manner and the second compression manner successively, where the fourth data is obtained by updating the third data; decompressing the first data and the second data in a first decompression manner corresponding to the second compression manner to obtain first intermediate data and second intermediate data; determining first differential data between the second intermediate data and the first intermediate data; and updating the first data based on the first differential data and the first intermediate data.

Abstract: The system includes a hub device, a first electronic device, and a second electronic device. The first electronic device includes a first ultra-wide band module and a millimeter-wave radar module. Position information of the second electronic device and position information of a user are obtained based on position measurement for the second electronic device, measurement for a human body position, and conversion that are performed by the first electronic device. At least one second electronic device automatically performs a preset operation based on the position information of the user and the position information of the second electronic device.

Abstract: The electronic device includes an ultra-wideband (UWB) chip and n antennas configured to construct a carrier antenna coordinate system, where n?3. A coordinate origin of the carrier antenna coordinate system is located at a location of an antenna 0 in the n antennas, an antenna i is located on a coordinate axis of the carrier antenna coordinate system, and a distance between the antenna i and the antenna 0 is less than or equal to ?, where ? is a wavelength of a UWB signal, and a value of i is sequentially obtained in {1, 2, . . . , n?1}. The electronic device receives UWB signals from m UWB base stations by using the n antennas, where m?3, and m is an integer. Spatial posture information of the electronic device is determined based on the received UWB signals.

Abstract: A method is implemented by an electronic device to mark a position of a home device. The method includes that the electronic device receives a first operation of a user, where the first operation is used to trigger the electronic device to mark a position of a first home device in a plurality of home devices. In response to the first operation, the electronic device obtains, by using a UWB chip, spatial position-attitude information of the electronic device when the electronic device separately points to the first home device in n orientations, and calculates spatial position information of the first home device based on the spatial position-attitude information of the electronic device in the n orientations.

Abstract: A technique for reconstructing a three-dimensional scene from monocular video adaptively allocates an explicit sparse-dense voxel grid with dense voxel blocks around surfaces in the scene and sparse voxel blocks further from the surfaces. In contrast to conventional systems, the two-level voxel grid can be efficiently queried and sampled. In an embodiment, the scene surface geometry is represented as a signed distance field (SDF). Representation of the scene surface geometry can be extended to multi-modal data such as semantic labels and color. Because properties stored in the sparse-dense voxel grid structure are differentiable, the scene surface geometry can be optimized via differentiable volume rendering.

Abstract: A photodetector includes: a first conductive type semiconductor layer; a semiconductor light absorption layer provided on the first conductive type semiconductor layer; a scatterer that is provided with a width equal to or less than a wavelength of incident light so as to be in contact with the semiconductor light absorption layer and forms a localized non-uniform electric field inside the semiconductor light absorption layer by scattering the incident light; a second conductive type semiconductor layer provided on the semiconductor light absorption layer so as to be apart from the scatterer; and an extraction electrode that is provided on the second conductive type semiconductor layer so as to be apart from the scatterer and extracts a photocurrent generated in the semiconductor light absorption layer due to formation of the localized non-uniform electric field.

Abstract: This application relates to a UWB base station anomaly detection method. The method includes: obtaining a first distance that is between each of the N UWB base stations and the UWB tag and that is obtained through measurement; obtaining an estimated value of spatial coordinates of the UWB tag in the first coordinate system based on the first distance and spatial coordinates of each UWB base station in the first coordinate system; obtaining a second distance between each UWB base station and the UWB tag based on the estimated value and the spatial coordinates of each UWB base station in the first coordinate system; and determining, based on the first distance and the second distance that correspond to each UWB base station, whether a UWB base station in the N UWB base stations is abnormal.

Abstract: An aging apparatus is connected to a drive circuit of a chip, and the drive circuit includes an inverter and at least one selection transistor. The apparatus is connected to an input end of the inverter and a gate of the selection transistor, an output end of the inverter is connected to a drain of the selection transistor, a source of the selection transistor and a second power supply end of the inverter are connected to a low voltage, and a first power supply end of the inverter is connected to a high voltage.

Abstract: A recombinant vector for expressing an immunogenic protein of porcine epidemic diarrhea virus (PEDV), a recombinant strain including the same and use thereof are provided. The recombinant vector for expressing the immunogenic protein S1B of the PEDV is provided, and the recombinant vector can be used to construct a recombinant Lactococcus lactis strain expressing the immunogenic protein S1B of the PEDV. The recombinant Lactococcus lactis strain and an expressed antigen protein S1B of the PEDV are mixed to prepare a microbial agent capable of preventing PEDV infection. The created microbial preparation may safely, effectively, and quickly prevent the PEDV infection in pigs and does not involve an immunization process.

Abstract: A dual-promoter recombinant vector expressing an immunogenic protein of an African swine fever virus (ASFV), a recombinant bacterium, and use thereof are provided. A recombinant gene consisting of a PpepN promoter, a coding sequence of a Usp45 secretion signal peptide, and a coding sequence of the immunogenic protein of the ASFV that are ligated in sequence is inserted downstream of a promoter in an original vector to obtain the dual-promoter recombinant vector expressing the immunogenic protein of the ASFV.

Abstract: Provided is a semiconductor device including: a semiconductor layer having an uneven structure configured to include a recessed portion on one surface side thereof; a first electrode film (first deposited film) provided on the one surface of the semiconductor layer; and a second electrode film (second deposited film) provided on a bottom surface of the recessed portion, wherein an enlarged portion having a cross-sectional area enlarged with respect to a portion on an opening portion side of the recessed portion is provided.

Abstract: An end effector for a torpedo car capping robot and a capping method thereof. which effector and method belong to the field of robot control. The end effector at least comprises: a pickup and release unit, a buffer unit, a heat preservation cover distance measurement unit, an end effector structure protection unit and a pneumatic execution unit.

Abstract: In accordance with an embodiment, a first electronic device is configured to communicate with a first ultra-wideband base station via ultra-wideband wireless communication and receive coordinates of the first electronic device that are based on a relative location between the first electronic device and the first ultra-wideband base station. The first electronic device is further configured to display a control window of a second electronic device in response to detecting that the first electronic device is located in a front area of the second electronic device and points to the second electronic device.

Abstract: A method for automatic control of a simulation error in an analog circuit and a use thereof are provided. The method first determines a measurement object and acquires a simulation standard by simulating the measurement object. Then, the method performs simulation based on different parameters of a circuit netlist to acquire an electrical parameter and/or circuit performance evaluation indicator of the measurement object during a simulation process. The method forms a simulation sampling dataset based on a simulation option combination parameter and an error corresponding to the simulation standard. Finally, the method optimizes and analyzes the simulation sampling dataset based on an artificial intelligence algorithm to acquire a simulation option combination that meets a simulation error requirement and takes a least time. The solution achieves automatic selection of the simulation option combination, improving the accuracy of the simulation result.

Abstract: An online detection device underwater elements includes an LIBS system in a sealing pressure chamber and an external airflow control system. The airflow control system has a gas probe bin and a gas source. An opening is formed at one end of the gas probe bin while the other end and the sealing pressure chamber are hermetically partitioned through a glass window. A laser in the LIES system outputs laser to an underwater object surface to be detected for generating plasma spectra. A spectrometer collects plasma spectra returned along an original optical path. When the device operates in water, the balance gas storage tank produces gas with the same pressure as underwater. A flow model is invoked according to the current water pressure to accurately control the air flow rate to form a stable gas environment in the gas probe, which improves the plasma excitation and collection efficiency.

Abstract: A deep learning-based video editing method can allow for automated editing of a video, reducing or eliminating user input, saving time and labor investments, and thereby improving video editing efficiency. Attribute recognition is performed on an object in a target video using a deep learning model. A target object is selected that satisfies an editing requirement of the target video. A plurality of groups of pictures associated with the target object from the target video are obtained using editing. An edited video corresponding to the target video is generated using the plurality of groups of pictures.