Abstract: Devices, systems, and methods for digital video coding, which includes matrix-based intra prediction methods for video coding, are described. In a representative aspect, a method for video processing includes determining, for encoding a current video block of a video into a bitstream representation of the current video block, a type of neighboring samples of the current video block to be used according to a rule, and adding, based on the determining, an encoded representation of the current video block to the bitstream representation, where the current video block is encoded using a matrix based intra prediction (MIP) mode in which a prediction block of the current video block is determined by performing, on previously coded samples of the video, a boundary downsampling operation, followed by a matrix vector multiplication operation, and selectively followed by an upsampling operation.

Abstract: A telescopic power supply track has a strip-shaped power supply base and a power taking module, the power taking module has a rod-shaped housing, a second external terminal, and two electrical contacts provided on the rod-shaped housing away from the second external terminal and are respectively set against the two conductive strips to position and electrically connect the power taking module; and two strip-shaped conductive members, arranged in the hollow structure of the rod-shaped housing, and the two ends of the strip-shaped conductive member are respectively electrically connected to the second external terminal and the two electrical contacts. The invention solves the problem of the long lead in the shelf in the prior art and the need to reserve the lead, reduces the time of the short line, increases beauty and improves safety.

Abstract: A centralized RAN architecture that includes lower layer transmission nodes (referred to as distributed units (DUs) herein) that connect to an upper layer RAN node (referred to as a centralized unit (CU) herein). A centralized RAN architecture may include a functional split, between the CU and DU, corresponding to the Packet Data Convergence Protocol (PDCP) and Radio Link Control (RLC) layers. In one embodiment, a CU may store copies of downlink PDCP PDUs that are transmitted to a first DU. The CU may retransmit, to a second DU, those of the PDCP PDUs which were previously transmitted to the source DU but have not been acknowledged as successfully delivered by the source DU.

Abstract: An identification of an item that was misclassified by a classification model constructed in accordance with a machine learning technique is received. One example of such a machine learning technique is a random forest. A subset of training data, previously used to construct the model, and that is associated with the misclassified item is identified. At least a portion of the identified subset is provided as output.

Abstract: Devices, systems and methods for digital video coding, which includes matrix-based intra prediction methods for video coding, are described. In a representative aspect, a method for video processing includes performing a conversion between a current video block of a video and a bitstream representation of the current video block according to a rule, where the rule specifies a relationship between samples of the current video block and matrices or offset values applied in a matrix weighted intra prediction (MIP) mode during the conversion, and where the MIP mode includes determining a prediction block of the current video block by performing, on previously coded samples of the video, a boundary downsampling operation, followed by a matrix vector multiplication operation, and selectively followed by an upsampling operation.

Abstract: Aspects of the disclosure provide a semiconductor memory device. The semiconductor memory device includes a memory cell array and peripheral circuitry coupled with the memory cell array. The memory cell array includes a plurality of memory cells. The peripheral circuitry includes programmable logic circuit that is configured, after the semiconductor memory device is powered on, to perform logic functions.

Abstract: A face recognition method and apparatus, and a storage medium are provided. The method includes: performing attribute feature extraction on an image to be processed including a target object to obtain N attribute features of the target object, N being an integer greater than 1; performing attention feature extraction on the image to be processed based on an attention mechanism to obtain N attention features of the target object; clustering the N attention features to obtain M clustered attention features, M being a positive integer and M<N; and determining the face recognition result of the target object according to the N attribute features and the M clustered attention features.

Abstract: Certain aspects of the present disclosure provide techniques for transmitting and processing reference signals, such as DMRS, that may account for mobility characteristics (e.g., that relate to a Doppler measurement) of a wireless node (e.g., a UE), such as Doppler measurements indicating how fast such a device is moving.

Abstract: A process for enhanced anaerobic digestion of sludge by alcoholization of food waste, comprising steps of: (1) ethanol prefermentation of food waste—feeding the food waste having been crushed into a uniform grain size into an ethanol production reactor, where the food waste is subjected to rapid and efficient anaerobic fermentation activated by addition of yeast and pH adjustment to produce ethanol and obtain a fermentation broth and a fermentation residue; (2) sludge pretreatment—pretreating the sludge after mixing with the fermentation residue; and (3) methanogenesis of anaerobic digestion—subjecting products resulting from the sludge pretreatment to be mixed with the fermentation broth and then to methanogenesis of anaerobic digestion in a methane production reactor. With this process, the sludge and the food waste can be treated efficiently via multi-phase reactions, and the resource recovery and biogas quality can be enhanced.

Abstract: The present invention provides a trench gate structure and a method of forming the same. The method comprises steps of forming a first trench on the surface of a substrate, a surface of a bottom of the first trench comprising a crystal face belonging to the first family of crystal faces, and a surface of a sidewall of the first trench comprising another crystal face belonging to a second family of crystal faces. With a face-selective wet etching, a specific crystal face is presented on the surface of the bottom of the trench and a thicker gate oxide layer is formed thereon after performing thermal oxidation to avoid from failure due to thinner gate oxide layer on the surface of the bottom, increase breakdown voltage, and improve reliability of the device.

Abstract: This application provides a communication method, a communications device, and a network device. When a beam failure occurs in a secondary cell of the communications device, the communications device sends a beam failure recovery request to the network device by using a PUCCH or a MAC CE, and determines an active state time of DRX based on a status of the beam failure recovery request. The communications device may also determine a stopping time of an uplink retransmission timer based on a running status of a configured grant timer, where the DRX of the communications device is in an active state when the uplink retransmission timer works; and monitors a PDCCH when the DRX of the communications device is in the active state.

Abstract: This application discloses a roaming charging method and apparatus. A blockchain network stores information related to roaming fee settlement, to reduce a possibility that the information related to roaming fee settlement is tampered with. In addition, a digital signature algorithm is used to sign roaming service statistics to generate an abstract, to further prevent the roaming service statistics from being tampered with. In this way, when a home carrier verifies a roaming account balance by using the roaming service statistics and a roaming charging rule, accuracy of a verification result can be ensured, and the home carrier can obtain an accurate roaming fee of a user.

Abstract: The present disclosure provides fusion polypeptides comprising GLP-1 and GDF15, the polypeptide complexes, and the conjugates thereof. Pharmaceutical compositions comprising the same and methods of treating diseases are also provided.

Abstract: The present invention belongs to the bioengineering field, and relates to a method for fermentation production of L-theanine by using an Escherichia coli genetically engineered bacterium. The engineered bacterium is obtained by serving a strain as an original strain, wherein the strain is obtained after performing a single copy of T7RNAP, a dual copy of gmas, xylR knockout, and sucCD knockout on an Escherichia coli W3110 genome, and by integrating genes xfp, pta, acs, gltA, and ppc, and knocking out ackA on the genome. The present invention has a high yield, and stable production performance; after 20-25 h, L-theanine has a titer of 75-80 g/L, and the yield is up to 52-55%. The fermentation broth is purified by membrane separation in combination with a cation-anion resin series technique. Moreover, the one-step crystallization yield is 72.3% and the L-theanine final product has a purity of 99%.