Abstract: Systems and methods for calculating load direction even under adverse environmental conditions are provided. A system may include sensing circuitry and processing circuitry. The sensing circuitry may sense a first parameter and a second parameter of the electrical waveform on the transmission line of the electric power distribution system. The processing circuitry may determine a present load direction of an electrical waveform using a first method based at least in part on the first parameter in response to detecting that the sensing circuitry is experiencing a first environmental condition. The processing circuitry may determine the present load direction of the electrical waveform using a second method based at least in part on the second parameter and not the first parameter in response to detecting that the sensing circuitry is experiencing a second environmental condition.

Abstract: Systems and apparatuses for generating customized user output are provided. The system may collect sensor data, associated with the user, from a variety of sources. The system may use the sensor data to generate a customized user output. The system may analyze the sensor data, and determine, based on the sensor data and the customized user output, one or more user recommendation outputs. The system may update the customized user output based on additional or subsequent sensor data, and/or based on whether or not the user recommendation output was completed, as determined from subsequent sensor data.

Abstract: An apparatus to facilitate workload scheduling is disclosed. The apparatus includes one or more clients, one or more processing units to processes workloads received from the one or more clients, including hardware resources and scheduling logic to schedule direct access of the hardware resources to the one or more clients to process the workloads.

Abstract: The present invention discloses a driver circuit and a driving method thereof including a first thin film transistor, a second thin film transistor, and a third thin film transistor. Increasing a photocurrent of the second thin film transistor, i.e., amplifying the photocurrent of the thin film transistor to a photosensitive thin film transistor, advantages enhancement of a signal intensity and a signal-noise ratio of the photocurrent read out by the read line to solve the issue of weak a photocurrent signal from the photosensitive display.

Abstract: A power supply ECU is configured to carry out extreme value search control for searching for a frequency at which a detected value of power loss is minimized, by oscillating a frequency of output power from an inverter. Then, the power supply ECU determines whether or not extreme value search in extreme value search control is poorly proceeding, and when extreme value search is poorly proceeding, the power supply ECU increases an amplitude of frequency oscillation of output power from the inverter.

Abstract: A battery backup system includes a battery pack, a charger, a discharger, a voltage detection and control circuit, a first protection switch, a second protection switch and a third protection switch. The battery pack provides a backup power to a DC bus. The charger, connected to the battery pack, receives an external power. The discharger has an input end and an output end. The voltage detection and control circuit detects whether the voltage of the DC bus is greater than first or second voltage drop point. The first, the second and the third protection switches respectively are connected between the battery pack and the input end, between the DC bus and the output end, and between the charger and the input end. The first protection switch is turned on/off according to whether the voltage of the DC bus is less than or greater than the first voltage drop point.

Abstract: Embodiments for self-managed adaptable models for prediction systems by one or more processors. One or more adaptive models may be applied to data streams from a plurality of data sources according to one or more data recipes such that the one or more adaptive models predict a plurality of target variables.

Abstract: A arrangement is disclosed for an on-chip system having an increased resolution for supply voltage measurements. The system includes a phase locked loop (PLL), a divider, and a timer. The PLL is configured to generate an oscillator signal. The divider is configured to divide the oscillator signal to generate an divided clock signal. The timer is configured to generate an application start signal and an analog to digital converter (ADC) start signal based on the oscillator signal and a timer delay (Tdelay). The timer delay (Tdelay) is based on the application start signal and the ADC start signal.

Abstract: Optimized memory usage and management is crucial to the overall performance of a neural network (NN) or deep neural network (DNN) computing environment. Using various characteristics of the input data dimension, an apportionment sequence is calculated for the input data to be processed by the NN or DNN that optimizes the efficient use of the local and external memory components. The apportionment sequence can describe how to parcel the input data (and its associated processing parameters—e.g., processing weights) into one or more portions as well as how such portions of input data (and its associated processing parameters) are passed between the local memory, external memory, and processing unit components of the NN or DNN. Additionally, the apportionment sequence can include instructions to store generated output data in the local and/or external memory components so as to optimize the efficient use of the local and/or external memory components.

Abstract: Both before and after a surprise clock stop, the apparatus and method of various embodiments supplies a stable and continuous clock to a memory module with a unique arrangement of circuit components, including a clock detector circuit, a clock-smoothing circuit, and one or more PLLs. Upon detection of a stopped host clock, a first PLL seamlessly switches to an alternate reference clock from an on-board crystal oscillator. A clock smoothing circuit allows the first PLL to maintain a steady phase and frequency without inducing glitches or period excursions greater than the natural jitter of the locked PLL; one or more optional downstream PLLs may drive additional clock domains.

Abstract: A current measurement circuit includes first, second, and third conductors, a first current sensor, a second current sensor, and current computation circuitry. The first conductor is configured to conduct a first phase current of a three-phase current. The second conductor is configured to conduct a second phase current of the three-phase current. The third conductor is configured to conduct a third phase current of the three-phase current. The first current sensor is coupled to the first, the second, and the third conductors. The second current sensor is coupled to the second conductor and the third conductor. The current computation circuitry is coupled to the first current sensor and the second current sensor, and is configured to determine the first current, the second current, and the third current by applying an inverse Clarke transform to the output of the first current sensor and the output of the second current sensor.

Abstract: An autoranging ammeter with fast dynamic response allows improved dynamic measurement of rapidly changing direct electrical currents. The ammeter utilizes a low-cost dual threshold comparator mechanism coupled with an analog-to-digital converter and digital processing to rapidly select the appropriate current shunt resistor.

Abstract: To obtain an impact wave by generation of arc discharge between a high-voltage-side electrode 31 connected to a high-voltage-side terminal of a pulse power generating device and a low-voltage-side electrode 32 grounded or connected to a low-voltage-side terminal of the power source. One of the high-voltage-side electrode 31 or the low-voltage-side electrode 32 is an annular electrode formed in an annular shape, the other electrode is a core electrode arranged inside the annular electrode, and arc discharge is generated between an inner peripheral portion of the annular electrode and an outer peripheral portion of the core electrode.

Abstract: A vehicle includes a lever; a magnetic sensor configured to detect an angle of movement of the lever; an inputter configured to receive setting information regarding a lane change and a drive mode from a user; and a controller configured to determine whether to lock the lever based on at least one of the detected angle, the setting information, and the drive mode, and unlock the lever based on the drive mode and the lane change.

Abstract: A keyswitch includes a circuit board, a bottom board abutting against the circuit board, a cap, a light receiver and a light emitter disposed on the circuit board and opposite to each other, a base disposed on the circuit board, a cover disposed on the base, a first elastic member disposed through the cover and the base, a support device, and a second elastic member fixed to the base and having a flexible rod. The support device is movably connected to the cap and bottom board and has a sheet structure. When the cap is pressed, the sheet structure blocks light transmission between the light emitter and the light receiver to generate a triggering signal. When the cap is pressed to make the flexible rod deform downward with the sheet structure and then cross the sheet structure, the flexible rod is released to collide with the cover to make sound.

Abstract: The present disclosure includes apparatuses and methods for operating neural networks. An example apparatus includes a plurality of neural networks, wherein the plurality of neural networks are configured to receive a particular portion of data and wherein each of the plurality of neural networks are configured to operate on the particular portion of data during a particular time period to make a determination regarding a characteristic of the particular portion of data.

Abstract: Apparatuses and methods of manufacturing same, systems, and methods for training deep learning machines are described. In one aspect, candidate units, such as detection bounding boxes in images or phones of an input audio feature, are classified using soft labelling, where at least label has a range of possible values between 0 and 1 based, in the case of images, on the overlap of a detection bounding box and one or more ground-truth bounding boxes for one or more classes.

Abstract: The present disclosure includes apparatuses and methods for operating neural networks. An example apparatus includes a plurality of neural networks, wherein the plurality of neural networks are configured to receive a particular portion of data and wherein each of the plurality of neural networks are configured to operate on the particular portion of data during a particular time period to make a determination regarding a characteristic of the particular portion of data.

Abstract: A signal processing circuit for dynamically processing an analog signal. The signal processing circuit may include a detection module and a fluctuation filter module. The detection module may be configured to detect a voltage of a first sub-signal of a first frequency band of the analog signal in real time and transmit a first control signal when the voltage of the first sub-signal is out of a first preset voltage range. The fluctuation filter module may be configured to remove a second sub-signal of a second frequency band of the analog signal from the analog signal upon receiving the first control signal. The second frequency band may cover the first frequency band.

Abstract: Embodiments of the present disclosure provide a method and apparatus for document recommendation by obtaining a plurality of first data for a source recommendation task from activities related to a source user using a plurality of first documents; obtaining a plurality of second data for a target recommendation task from activities related to a target user using a plurality of second documents; performing the target recommendation task based on the plurality of first data, the plurality of second data, and knowledge transferred from the source recommendation task to obtain a target recommendation model; and conducting document recommendation to the target user using the target recommendation model.