Abstract: An outer loop initial value adjustment method, a device, and a readable storage medium are disclosed. The method may include: acquiring feature information of a plurality of User Equipment (UEs) scheduled by a cell; performing raster division on the feature information of the plurality of UEs to obtain a plurality of rasters; determining an outer loop corresponding to each of the rasters, where an outer loop value of each raster is within a target Block Error Rate (BLER) range; determining a target raster corresponding to a target UE according to the feature information of the target UE, where the target UE is one of the plurality of UEs scheduled by the cell; and determining the outer loop value corresponding to the target raster as an initial outer loop value of the target UE.

Abstract: The present invention discloses potassium sodium bismuth niobate tantalate zirconate ferrite ceramics with non-stoichiometric Nb5+ and a preparation method therefor. A ceramic powder with a general formula of (K0.45936Na0.51764Bi0.023)(Nb0.89958+0.957xTa0.05742Zr0.04Fe0.003)O3 (?0.01?x?0.04) is prepared by a traditional solid phase method; and then piezoelectric ceramics are prepared by traditional electronic ceramic preparation processes such as granulating, molding, binder removal, sintering and silvering test. An excessive amount of Nb5+ doping improves the temperature stability of the ceramics by providing a domain wall pinning effect. This result demonstrates the promise of potassium sodium bismuth niobate tantalate zirconate ferrite ceramics for a wide range of applications, including sensors, actuators, and other electronic devices.

Abstract: An electrochromic device includes a charge balancing thin film comprised of a new vanadium oxide with a formula of VOx, which provides a high charge density, low coloration efficiency, an electroactive voltage in close proximity to those of some electrochromic materials, and high chemical and electrochromic stability. Vanadium oxide can be without doping or doped with others. The VOx charge balancing thin film has a porous nanostructure and is amorphous or a combination of amorphous and polycrystalline, and can work with electrochromic conjugated polymer in the device in a minimally color changing mode. A method to design a material for a charge balancing thin film to pair with a working electrode and obtain a low device voltage in an electrochromic device is disclosed. Methods to prepare related charge balancing thin films are also disclosed.

Abstract: A computer-implemented method includes determining, based on a configured computing resource share at a previous moment/time period based on an estimated traffic sequence of each application in a target period a representation vector of each application of a plurality of applications and a pre-estimated central processing unit utilization of each application. The pre-estimated central processing unit utilization of each application, the representation vector of each application, and the estimated traffic sequence of each application in the target period are input into a decision network and a current computing resource configuration policy of a first moment/time period based on an output result of the decision network.

Abstract: First graph structure data indicating connectivity relationships mined for N object nodes is generated based on node representation parameters. The first graph structure data and node traffic data including T traffic values of each object node at T time points is processed to obtain N first embedded representations of the N object nodes. Second graph structure data indicating original connectivity relationships pre-generated for N object nodes and the node traffic data are processed to obtain N second embedded representations of the N object nodes. Fusion processing is performed on the N first embedded representations and the N second embedded representations to obtain a fused representation matrix, which is processed with the T traffic values. A predicted traffic value of each object node after the T time points is determined. Parameters are updated in a traffic prediction system based on the predicted traffic value and a corresponding actual traffic value.

Abstract: A semiconductor device includes a substrate and a conductive pad coupled to the substrate. A first solder mask is coupled to the substrate and to a portion of the conductive pad so the first solder mask covers the portion of the conductive pad and extends above the conductive pad. A second solder mask is coupled to a portion of the first solder mask and extends above the first solder mask.

Abstract: This application discloses an automatic switching system and method for a front end processor (FEP). The system includes: at least one external device and a FEP assembly. The FEP assembly is connected to the at least one external device. The FEP assembly provides services upward by using a primary memory, a primary TO manager, a secondary memory, and a secondary TO manager, and is connected downward to the at least one external device by using at least one primary driver and at least one secondary driver. The FEP assembly is configured to use the at least one secondary driver as a new primary driver when there is a fault in a communication link between the at least one primary driver and the at least one external device, to transmit a control instruction to the at least one external device and acquire data from the at least one external device.

Abstract: Provided are a method and an apparatus for adjusting an inner loop value, The method includes: repeatedly executing the following steps, wherein a history inner loop adjustment factor of each time-domain scheduling unit in a radio frame is initialized to be an initial inner loop adjustment factor of each time-domain scheduling unit: determining a block error rate corresponding to each time-domain scheduling unit according to ACK/NACK information corresponding to each time-domain scheduling unit, wherein the radio frame comprises N time-domain scheduling units, N being a positive integer; determining a current inner loop adjustment factor of each time-domain scheduling unit according to the block error rate corresponding to each time-domain scheduling unit and the history inner loop adjustment factor of each time-domain scheduling unit; and adjusting an inner loop value corresponding to each time-domain scheduling unit according to the current inner loop adjustment factor of each time-domain scheduling unit.

Abstract: Implementations of the present specification provide a method of generating a regular expression for a parameter of an application system and a method and an apparatus of checking validity of a parameter of an application system. To generate a regular expression for a parameter of the application system, parameter values of each parameter in each method code block are obtained from a historical implementation of the application system. Common information is extracted from the corresponding parameter values for each parameter in each method code block. Then a corresponding regular expression is generated for each parameter in each method code block based on the extracted common information.

Abstract: Techniques related to implementing convolutional neural networks for face or other object recognition are discussed. Such techniques may include applying, in turn, a depth-wise separable convolution, a condense point-wise convolution, and an expansion point-wise convolution to input feature maps to generate output feature maps such that the output from the expansion point-wise convolution has more channels than the output from the condense point-wise convolution.

Abstract: The present application relates to the technical field of communications, and in particular to a method for scheduling spectrum resources and a storage medium. The method for scheduling spectrum resources includes: obtaining a grid according to dividing a cell in a network, each grid corresponding to one resource block or one resource block group; obtaining offline feature data; performing an interference mark on the grid according to the offline feature data, to obtain a mark model; and scheduling spectrum resources according to the mark model.

Abstract: Techniques and apparatus for generating dense natural language descriptions for video content are described. In one embodiment, for example, an apparatus may include at least one memory and logic, at least a portion of the logic comprised in hardware coupled to the at least one memory, the logic to receive a source video comprising a plurality of frames, determine a plurality of regions for each of the plurality of frames, generate at least one region-sequence connecting the determined plurality of regions, apply a language model to the at least one region-sequence to generate description information comprising a description of at least a portion of content of the source video. Other embodiments are described and claimed.

Abstract: An aliphatic copolycarbonate and a preparation method thereof are disclosed. The aliphatic copolycarbonate contains a structural unit represented by formula (1) and a structural unit represented by formula (2): where R1 is a C3-C10 alkylene group, and R2 is an alicyclic group. In the present disclosure, a specific alicyclic monomer is introduced into the aliphatic polycarbonate molecular chain to obtain a novel aliphatic copolycarbonate. The polymer in the present disclosure has properties such as relatively high melting point, glass transition temperature and thermal stability, and has better thermal properties.

Abstract: A method and apparatus for determining an outer loop value, a device, and a storage medium are disclosed. The method may include: determining a pre-trained outer loop initialization model based on current feature data of a user equipment (S11); and determining an initialized outer loop value of the user equipment based on a current air interface measurement value of the user equipment and the outer loop initialization model (S12).

Abstract: An example apparatus for semantic image segmentation includes a receiver to receive an image to be segmented. The apparatus also includes a gated dense pyramid network including a plurality of gated dense pyramid (GDP) blocks to be trained to generate semantic labels for respective pixels in the received image. The apparatus further includes a generator to generate a segmented image based on the generated semantic labels.

Abstract: An evaporative device has a piece of at least partially-delignified plant material. The at least partially-delignified plant material has a modified microstructure including a plurality of vessels, a plurality of fibers, and a plurality of engineered micropores. Each vessel can define a first lumen having a maximum cross-sectional dimension of at least 100 ?m. Each fiber can define a second lumen having a maximum cross-sectional dimension less than or equal to 20 ?m. The engineered micropores can extend through walls of the vessels or fibers so as to fluidically interconnect the first and second lumina. In some embodiments, the plant material is reed or bamboo.

Abstract: One embodiment provides for a graphics processor comprising a cache memory and a graphics core coupled with the cache memory. The graphics core includes circuitry configured to generate an approximate weight matrix including a set of one-hot coded weights, perform a forward compute pass with mini batch samples to compute a loss function, perform a backward compute pass to compute a gradient update via stochastic gradient descent according to a loss update, and update the approximate weight matrix based on the gradient update to generate an updated weight matrix.

Abstract: An example apparatus for visual question answering includes a receiver to receive an input image and a question. The apparatus also includes an encoder to encode the input image and the question into a query representation including visual attention features. The apparatus includes a knowledge spotter to retrieve a knowledge entry from a visual knowledge base pre-built on a set of question-answer pairs. The apparatus further includes a joint embedder to jointly embed the visual attention features and the knowledge entry to generate visual-knowledge features. The apparatus also further includes an answer generator to generate an answer based on the query representation and the visual-knowledge features.

Abstract: A piece of natural plant material is subjected to one or more chemical treatments to remove substantially all lignin therefrom, thereby allowing the extraction of delignified, cellulose-based fibers. For example, the natural plant material can be a grass, such as bamboo or gladiolus. Subsequent drying of the extracted fiber densifies the structure, yielding improved mechanical properties. In some embodiments, the extracted fibers can be used, either alone or in combination with other materials, as a structural material. For example, the extracted fibers can be embedded within, infiltrated with, coated by, or otherwise combined with a polymer or concrete to form a composite material.