Abstract: A system and method for loading data into column-partitioned database tables. The system and method incorporate a mechanism for buffering data extracted from the rows of a source table in column-oriented fashion within an in-buffer memory, enabling an efficient bulk-write of large arrays of values from the buffer into column-partitioned database tables. The system and method may also include optimizations for grouping columns according to data types and altering the order in which columns are inserted into the database tables.

Abstract: An HARQ retransmission method and apparatus and a transmission device are provided. The method includes: determining a transmission parameter; and selecting dynamically a transmission starting position of each retransmission redundancy version from a circular buffer according to the transmission parameter.

Abstract: A method for enabling light detection and ranging (LiDAR) scanning is provided. The method is performed by a system disposed or included in a vehicle. The method comprises receiving a first laser signal. The first laser signal has a first wavelength. The method further includes generating a second laser signal based on the first laser signal. The second laser signal has a second wavelength. The method further includes providing a plurality of third laser signals based on the second laser signal; and delivering a corresponding third laser signal of the plurality of third laser signals to a respective LiDAR scanner of the plurality of LiDAR scanners. Each of the LiDAR scanners are disposed at a separate location of the vehicle such that each of the LiDAR scanners is capable of scanning a substantial different spatial range from another LiDAR scanner. LiDAR systems can use multi-wavelength to provide other benefits as well.

Abstract: Described herein are semi-solid polymer electrolytes (SSPEs) based on a polymer backbone incorporating a flame-retardant crosslinker and fluorinated counterions that are useful in the production of high energy rechargeable lithium metal batteries. The described SSPEs are not liquid electrolytes, are not solid state electrolytes (SSEs), and are differentiated from standard state-of-the-art gel polymer electrolytes (GPEs). The described SSPEs are formed from a first solvent, an optional second solvent, a crosslinker, a lithium salt, and an initiator. The unique coordination structure of the described SSPEs yields non-flammable, low-volatility, non-vaporizable, high Coulombic efficiency (CE), stable solid-electrolyte-interphase (SEI)-forming electrochemical devices, such as lithium metal rechargeable batteries, that are easily adaptable to existing mass-production lines.

Abstract: The disclosed technology involves autonomously identifying goals and sub-goals from a user utterance and generating responses to the user based on the goals and sub-goals.

Abstract: The invention relates to the technical field of device detection, and in particular provides a status detection method and device for a vehicle light signaling control device, and a vehicle, so as solve a technical problem of how to accurately and reliably detect an operating status of a vehicle light signaling control device. For this purpose, according to the method of an embodiment of the invention, a first resistance jump sequence of a currently detected position resistance may be obtained according to historical position resistance data and the currently detected position resistance; and then it is determined whether the vehicle light signaling control device has malfunctioned, according to a comparison result of the first resistance jump sequence and a preset second resistance jump sequence.

Abstract: Embodiments discussed herein refer to LiDAR systems and methods that enable substantially instantaneous power and frequency control over fiber lasers. The systems and methods can simultaneously control seed laser power and frequency and pump power and frequency to maintain relative constant ratios among each other to maintain a relatively constant excited state ion density of the fiber laser over time.

Abstract: Methods and systems are described for generating dynamic interface options using machine learning models. The dynamic interface options may be generated in real time and reflect the likely goals and/or intents of a user. The machine learning model may provide these features by interpreting multi-modal feature inputs. For example, the machine learning model may include a first machine learning model, wherein the first machine learning model comprises a convolutional neural network, and a second machine learning model, wherein the second machine learning model performs a Weight of Evidence (WOE) analysis.

Abstract: A low-dispersion single-mode optical fiber includes a core and claddings covering the core. The core has a relative refractive index difference of 0.30-0.65% and a radius of 2.5-4.5 ?m. The claddings include first, second, third cladding layers and an outer cladding arranged sequentially from inside to outside. The first cladding layer covers the core, and has a relative refractive index difference of ?0.70% to ?0.30% and a radius of 4.5-7.5 ?m. The second cladding layer covers the first cladding layer, and has a relative refractive index difference of ?0.20% to ?0.25% and a radius of 7.0-12.0 ?m. The third cladding layer covers the second cladding layer, and has a relative refractive index difference of ?0.60% to 0.00% and a radius of 10.0-20.0 ?m. The outer cladding covers the third cladding layer, and is a layer made of pure silicon dioxide glass.

Abstract: Methods and systems are described for generating dynamic interface options using machine learning models. The dynamic interface options may be generated in real time and reflect the likely goals and/or intents of a user. The machine learning model may provide these features by interpreting multi-modal feature inputs. For example, the machine learning model may include a first machine learning model, wherein the first machine learning model comprises a convolutional neural network, and a second machine learning model, wherein the second machine learning model performs a Weight of Evidence (WOE) analysis.

Abstract: A mobile device includes a sensing unit configured to determine whether the stylus pen is separated from the mobile device; a communication unit configured to receive a control signal including a first control signal and a second control signal from the stylus pen separated from the mobile device; and a controller configured to operate in a first control mode for performing an operation of an application being executed on the mobile device. The controller performs a first operation of the application in response to the first control signal, and performs a second operation of the application in response to the second control signal.

Abstract: A recombinant fusion protein comprising a human erythropoietin peptide portion linked to an immunoglobulin peptide portion is described. The fusion protein has a prolonged half-life in vivo in comparison to naturally occurring or recombinant native human erythropoietin. In one embodiment of the invention, the protein has a half-life in vivo at least three fold higher than native human erythropoietin. The fusion protein also exhibits enhanced erythropoietic bioactivity in comparison to native human erythropoietin. In one embodiment, the fusion protein comprises the complete peptide sequence of a human erythropoietin (EPO) molecule and the peptide sequence of an Fc fragment of human immunoglobulin IgG1. The Fc fragment in the fusion protein includes the hinge region, CH2 and CH3 domains of human immunoglobulin IgG1. The EPO molecule may be linked directly to the Fc fragment to avoid extraneous peptide linkers and lessen the risk of an immunogenic response when administered in vivo.

Abstract: Embodiments discussed herein refer to using LiDAR systems that uses a rotating polygon with a multi-facet mirror. Such multi-facet galvanometer mirror arrangements generate a point map that has reduced curvature.

Abstract: In some embodiments, a method generates a representation for a video that represents a relationship of a bitrate to a quality measurement for the video and selects candidate transcoding points on the representation. A plurality of candidate profile ladders are generated based on the candidate transcoding points. Each of the plurality of candidate profile ladders includes a different combination of profiles and a profile is associated with at least one of a bitrate and a quality measurement. The method selects a profile ladder from the plurality of candidate profile ladders based on analyzing the bitrate or the quality of the profile ladder and uses the profile ladder for the video.

Abstract: The present invention discloses a microencapsulation method for improving stability of anthocyanin, a product therefrom and use thereof. A preparation method of anthocyanin microcapsules includes: (1) taking sodium alginate as a wall material, adding sodium alginate and calcium carbonate into water, and swelling for 1-2 h to obtain a wall material gel system; (2) taking anthocyanin prepared by a special process as a core material, and fully and uniformly mixing the wall material gel system with an anthocyanin solution to obtain a water phase; (3) mixing Span80 and vegetable oil to obtain an oil phase, mixing the water phase with the oil phase, and magnetically stirring for emulsifying to obtain a W/O emulsion; and (4) adjusting the pH of the W/O emulsion to be acidic, mixing the W/O emulsion with a salt buffer solution, standing for 1-2 h, and then separating the oil phase and the water phase.

Abstract: This application provides a storage device and a data writing method. The storage device may be used in a neural network. The storage device includes a memristor unit, a current-controlled circuit, and a write circuit. The memristor unit has a structure of one-transistor and one-resistive random access memory 1T1R. The current-controlled circuit is configured to limit a current passing through the memristor unit to a target current, where the target current is determined based on target conductance of the memristor unit and a gate voltage of the transistor, and the target conductance is used to indicate target data to be written into the memristor unit. The write circuit is configured to load a write voltage to the memristor unit in cooperation with the current-controlled circuit, to write the target data to the memristor unit.

Abstract: A method includes obtaining a source training dataset that includes a plurality of source training images and obtaining a target training dataset that includes a plurality of target training images. For each source training image, the method includes translating, using the forward generator neural network G, the source training image to a respective translated target image according to current values of forward generator parameters. For each target training image, the method includes translating, using a backward generator neural network F, the target training image to a respective translated source image according to current values of backward generator parameters. The method also includes training the forward generator neural network G jointly with the backward generator neural network F by adjusting the current values of the forward generator parameters and the backward generator parameters to optimize an objective function.

Abstract: Systems and methods are provided herein for utilizing a knowledge base to improve online automated dialogue responses based on machine learning models. Contextual customer data stored in external memory may be used for retraining a machine learning model to incorporate new observations into the model and to reduce bias and/or improve fairness in associated automated responses without having to retrain an entire memory architecture. The disclosed technology may improve the accuracy of machine learning models by using potentially private contextual customer data to inform the model while eliminating the ability of an intruder to access such data when the model is utilized in cloud-based services.

Abstract: The invention provides a piezoelectric micro-electromechanical system (MEMS) microphone includes a base with a cavity and a piezoelectric diaphragm arranged on the base. The base has a ring base and a support column. The piezoelectric diaphragm includes a plurality of diaphragm sheets. Each diaphragm sheet has a fixing end connected with the support column and a free end suspended above the cavity. The widths of the diaphragm sheets are gradually increased from the fixing ends to the free ends. According to the piezoelectric MEMS microphone provided by the invention, under sound pressure, the free ends vibrate, wide free ends drive short fixing ends, and the diaphragm sheets close the fixing ends generate greater deformation to generate more charge. Therefore, the sensitivity can be further improved.