Abstract: Video processing with a multi-quality loop filter using a multi-task neural network is performed by at least one processor and includes generating model IDs, based on quantization parameters in an input, selecting a first set of masks, each mask in the first set of masks corresponding to one of the generated model IDs, performing convolution of first weights of a first set of neural network layers and the selected first set of masks to obtain first masked weights, and selecting a second set of neural network layers and second weights, based on the quantization parameters, generating a quantization parameter value, based on the quantization parameters, and computing an inference output, based on the first masked weights and the second weights, using the generated quantization parameter value.

Abstract: A wireless device receives radio resource control messages including configuration parameters of transmission configuration indication (TCI) states. A control element is received and includes an indicator, associated with a TCI state of the TCI states, of the TCI state being associated with a channel type of channel types that include: a first channel type including a downlink control channel and a downlink shared channel, and a second channel type including an uplink control channel and an uplink shared channel. A downlink control information is received that includes a TCI field indicating the TCI state. Based on the indicated TCI and the indicator of the TCI being associated with the channel type, the wireless device communicates with a base station via one or more channels of the channel type.

Abstract: A method of healthcare allocation includes receiving, at an artificial intelligence engine, health data from sensors over a period of time. The sensors read the health data of a plurality of patients. Thereby, as the artificial intelligence engine receives the health data, the artificial intelligence engine learns a baseline health assessment of each of the patients. After the period of time elapses, the artificial intelligence engine continuously receives the health data from the sensors. If the health data singularly or in combination indicates an absolute healthcare issue exists, the artificial intelligence engine allocates/recommends at least one resource to one of the patients that is associated with the health data. Periodically, the artificial intelligence engine scans all patients and generates a priority for help for each of the patients and then allocates/recommends at least one resource to each patient in a subset of the patients based upon the priority.

Abstract: A computing device includes signal generation circuitry and also includes a location on the computing device that is operative to couple a signal generated by the signal generation circuitry into a user. For example, the computing device includes signal generation circuitry that generates a signal that includes information corresponding to a user and/or an application that is operative within the computing device. The signal generation circuitry couples the signal into the user from a location on the computing device based on a bodily portion of the user being in contact with or within sufficient proximity to the location on the computing device that facilitates coupling of the signal into the user. Also, the signal may be coupled via the user to another computing device that includes a touchscreen display that is operative to detect and receive the signal.

Abstract: Techniques are described to automate deployment of an application as a cloud computing service. Such deployment may be accomplished on any underlying host infrastructure. As mentioned above, a typical application may engage in a number of build and deployment processes in order to run on a host computer with an infrastructure type. The techniques described herein may leverage an artificial intelligence computation library to identify which build and deployment processes are required by the application. Once identified, the user may select a host infrastructure type and create a container package with a runtime machine and components configure to run these build and deployment processes for that host infrastructure type. The container package may be used to generate an image from which the host computer can run the application as the cloud computing service. Other embodiments are described and claimed.

Abstract: A system and method of dynamically controlling a reservation of resources within a cluster environment to maximize a response time are disclosed. The method embodiment of the invention comprises receiving from a requestor a request for a reservation of resources in the cluster environment, reserving a first group of resources, evaluating resources within the cluster environment to determine if the response time can be improved and if the response time can be improved, then canceling the reservation for the first group of resources and reserving a second group of resources to process the request at the improved response time.

Abstract: An obstacle detection and notification system for a motorcycle. The system includes a forward looking camera and a backward looking camera mountable to the motorcycle and a processor in operable communication with the forward looking camera and the backward looking camera. The processor executes program instructions to execute processes including: receiving video from the of the forward looking camera and the backward looking camera, performing a computer vision and machine learning based object detection and tracking process to detect, classify and track obstacles in the video and to output detected object data, defining a blind spot region around one or more other vehicles using the detected object data, determining whether the motorcycle is located in the blind spot region, and outputting audible, tactile or visual feedback, via an output system, to a rider of the motorcycle when the motorcycle is determined to be located in the blind spot region.

Abstract: A decoder may receive, from a bitstream for a block, a residual block of samples, a first component, and a second component. The decoder may determine reference samples for intra prediction of the block. The decoder may generate a prediction of the block from the reference samples based on a ratio of the first component to the second component. The ratio determines a prediction angle. The decoder may generate a decoded block from the prediction of the block and the residual block.

Abstract: A power system includes a battery and a controller. The controller inhibits charge of the battery according to voltage or current values sensed before and after contactors electrically connected to the battery are commanded to open and indicating that a leakage resistance associated with one of the contactors increases after the one of contactors is commanded to open, and a duration of a continuous voltage drop across another of the contactors after the another of the contactors is commanded to open exceeds a threshold.

Abstract: A vehicle includes a two-panel support structure for each side of a bed of a pick-up truck. Each side includes an upper panel and a lower panel between the upper panel and the bed. The upper panel includes multiple indentations from a cab-side to a tailgate side of the bed. The lower panel includes a plurality of extensions that extend into one or more of the multiple indentations. The combination of each indentation and extension creates a pocket into which a platform, board, or other support structure, such as cross-brace, can be placed. The support structure can extend from one indentation on one side to a corresponding indentation on another side of the bed. The support structure creates support in conjunction with other support structures for a larger platform, such as a piece of plywood the size of the bed.

Abstract: A method is provided for controlling the operation of an offset well located near an active well that is undergoing a hydraulic fracturing operation that may produce a fracture driven interference (FDI) event to the offset well. The method includes providing an FDI intervention system that includes a computer-implemented predictive model for determining a risk of the FDI event occurring during the hydraulic fracturing operation, calculating a risk-weighted FDI event cost of the FDI event impacting production from the offset well, and calculating a defensive intervention implementation cost to apply a defensive intervention on the offset well to mitigate harm from an FDI event. The method includes calculating a cost comparison based on a comparison of the defensive intervention implementation cost and the risk-weighted FDI event cost. The method concludes with automatically controlling the operation of the offset well with the FDI intervention system based on the cost comparison.

Abstract: This document describes techniques and apparatuses for performing color calibration of display modules using a reduced number of display characteristic measurements. In aspects, methods include generating a measured lookup table (50) for a source display module (32), using dynamic optimization to down-sample the measured lookup table and select a set of color patches, sending the color patches to a testing display module; measuring output values for the testing display module; generating a sparse lookup table (52) relating the color patches and the measured output values, up-sampling the sparse lookup table to a forward lookup table (54), and inverting the forward lookup table to generate a reverse lookup table (56). The reverse lookup table (56) can be utilized to determine correct output values (29) for driving a target display module to generate color lights within a display module color gamut.

Abstract: A SiC MOSFET device with alternating p-well widths, including an undulating channel, is described. The undulating channel provides current paths of multiple widths, which enables optimization of on-resistance, transconductance, threshold voltage, and channel length. The multi-width p-well region further defines corresponding multi-width Junction FETs (JFETs). The multi-width JFETs enable improved response to a short-circuit event. A high breakdown voltage is obtained by distributing a high electric field in a JFET of a first width into a JFET of a second width.

Abstract: The technology relates to bandwidth constrained communication systems with neural network based detection. In some embodiments, a bandwidth constrained equalized transport (BCET) communication system comprises: a transmitter comprising an error control code encoder, a pulse-shaping filter, and a first interleaver; a communication channel; and a receiver comprising a neural network processing block that processes a received signal. The error control code encoder can append redundant information onto the signal. The pulse-shaping filter can intentionally introduce memory into the signal in the form of inter-symbol interference. The first interleaver can change a temporal order of the symbols in the signal. The error control code encoder can be a low-density parity-check (LDPC) error control code encoder. The neural network can be trained with positive mappings between transmitted and decoded training signals, or negative mappings between training signals and a null space of an LDPC generation matrix.

Abstract: A system and method for combining at least two similarly managed credit accounts into a single ONE card account is described. The system identifies a given cardholder having a plurality of similarly managed credit accounts. Generates a single ONE card account for the given cardholder. Incorporates the plurality of similarly managed credit accounts into the single ONE card account. Manages any future transactions for each of the plurality of similarly managed credit accounts via the single ONE card account. Provides, to the given cardholder, a single ONE card account statement, the single ONE card account statement containing information previously provided separately by each of the plurality of similarly managed credit accounts.

Abstract: A first wireless device receives, from a base station, configuration parameters for a resource allocation mode 1 of a sidelink bearer between the first wireless device and a second wireless device. The first wireless device receives, from the base station, at least one parameter indicating the sidelink bearer to transition from the resource allocation mode 1 to a resource allocation mode 2. The first wireless device transmits, to the second wireless device, transport blocks via the sidelink bearer based on the resource allocation mode 2.

Abstract: One or more implementations may include operations that include obtaining, by a first device, second device information related to a capability of a second device with respect to operating in a first frequency band. The operations may also include causing an adaptive radio of the first device to transition between operating in the first frequency band and operating in a second frequency band based on the second device information.

Abstract: A base station transmits, to a wireless device, at least one first radio resource control message comprising configuration parameters for a first session associated with a first session identifier between the wireless device and a user plane function and a second session associated with a second session identifier different from the first session identifier, for packet duplication of the first session, between the wireless device and the user plane function. The base station transmits, to the wireless device, at least one second radio resource control message indicating activation of the packet duplication. Based on the activation of the packet duplication, the base station receives original packets via the first session associated with the user plane function and duplicated packets of the first session via the second session.

Abstract: A method for uplink control multiplexing of a physical uplink control channel (PUCCH) repetition may include receiving, by a wireless device, a first message indicating PUCCHs for repetitions of uplink control information (UCI). The method may include receiving a second message indicating a number of repetitions of a transport block via physical uplink shared channels (PUSCHs) and indicating a beta offset for multiplexing the UCI via a PUSCH of the PUSCHs. The method may include determining a coding rate based on the beta offset and the number of repetitions. The method may include transmitting, based on the coding rate and in response to the PUCCHs overlapping with the PUSCHs in at least one symbol, the UCI and the transport block multiplexed in the PUSCHs.

Abstract: The disclosure provides a secondary light assembly, a night light assembly, a light fixture, and a method and unit for controlling the secondary light assembly. The secondary light assembly can be mounted on a main light assembly of the light fixture. The main light assembly is configured to provide first lighting light. The secondary light assembly includes: light emitting devices configured to provide second lighting light, a main sensor configured to detect the first lighting light of the main light assembly and generate a lighting light detection signal according to a detection result; and a control part configured to receive the lighting light detection signal from the main sensor and control the light emitting devices of the secondary light assembly at least based on the lighting light detection signal.