Abstract: A two-stage machine learning model is used to for categorization of a dataset, such as transactions. A plurality of complementary base machine learning models are used to generate initial inference results and associated measures of inference confidence from the dataset, which are collected as a meta dataset. Each of the complementary models is associated with a different part of the dataset in which it has a higher accuracy in that part than the other models. The meta dataset is provided as input to a meta machine learning model, which is trained to produce a final inference result, and a confidence score model, which is trained to produce a confidence score associated with the final inference result.

Abstract: The disclosure provides a method for monitoring a short-term voltage stability of a power system. The method includes: obtaining a topology and time series of a current power system; inputting the topology and the time series of the current power system into a trained spatial-temporal graph network model, the trained spatial-temporal graph network model being obtained by classification learning on a spatial-temporal graph network model based on a simulation sample dataset; outputting a status of the short-term voltage stability of the power system; and sending an alarm signal in response to outputting an unstable status of the short-term voltage stability of the power system.

Abstract: Embodiments of the disclosure provide an optical sensing system and an optical sensing method thereof. The optical sensing system comprises a light source configured to emit an optical signal into an environment surrounding the optical sensing system. The optical sensing system further comprises a photodetector configured to receive the optical signal reflected from the environment of the optical sensing system, and convert the optical signal to an electrical signal, where the photodetector is disposed in a package. The optical sensing system additionally comprises a readout circuit configured to generate a readout signal based on the electrical signal received from the photodetector, where the readout circuit is disposed in the same package as the photodetector and connected to the photodetector through routings in the package.

Abstract: An excess heat-generating element is coupled to a heat sink through a heat conduction path. A thermal switch is mounted in the heat conduction path. A temperature-sensitive element is coupled to the heat conduction path on a same side of the thermal switch as the excess heat-generating element. A temperature monitor is mounted adjacent the temperature-sensitive element. A temperature controller has an input coupled to the temperature output of the temperature monitor and an output control line coupled to an input of the thermal switch. The temperature controller switches off the thermal switch, in response to detecting a temperature below a temperature threshold from the temperature output. When the thermal switch it off, it impedes heat flow from the excess heat-generating element to the heat sink, and the heat flow is redirected to increase heat flow from the excess heat-generating element to the heat-sensitive element.

Abstract: Devices and techniques are generally described for determining named entity recognition tags. In various examples, first input data representing a natural language input may be determined. In some examples, a first machine learned model may determine first data comprising a first encoded representation of the first input data. In various examples, second data representing a grouping of text of the first input data may be determined based at least in part on the first data. In some examples, first entity data may be determined by searching a memory layer using the second data. In at least some examples, the first entity data and the first data may be combined to generate third data. In various examples, output data comprising a predicted named entity recognition tag may be generated for the grouping of text based at least in part on the third data.

Abstract: Embodiments of the disclosure provide methods for microfabricating an omni-view peripheral scanning system. One exemplary method may include separately fabricating a reflector and a scanning MEMS mirror, and then bonding the microfabricated reflector with the scanning MEMS mirror to form the omni-view peripheral scanning system. The microfabricated reflector may include a cone-shaped bottom portion, and a via hole across the cone-shaped bottom portion. The microfabricated scanning MEMS mirror may include a MEMS actuation platform and a scanning mirror supported by the MEMS actuation platform. The scanning MEMS mirror may face the cone-shaped bottom portion of the reflector when forming the omni-view peripheral scanning system.

Abstract: An orthopedic surgical robot system; the orthopedic surgical robot system comprises: a robotic arm (10), a distal end (170) of the robotic arm (10) being used for providing a locked working point (O) for the orthopedic surgical robot system; a multi-link mechanism (20), wherein the multi-link mechanism (20) comprises: a first link (210), the first link (210) being coupled to the distal end (170) of the robotic arm (10); and a second link (220), a proximal end of the second link (220) being connected to the first link (210) in a movable manner, so that the movement of the second link (220) relative to the first link (210) provides the orthopedic surgical robot system with a locked working plane relative to the locked working point (O); a bone saw (30), the bone saw being coupled to the distal end of the second link (220) and operable in the locked working plane; and a control system (40), the control system (40) being coupled to the robotic arm (10), and controlling, on the basis of a basic control amount of t

Abstract: Provided is a display panel. The display panel includes: a substrate, at least one row of light-emitting elements disposed on the substrate, a plurality of pixel drive circuits, a plurality of connection lines, and a plurality of compensation portions. A length of a first connection line electrically connected to a first pixel drive circuit is less than a length of a second connection line electrically connected to a second pixel drive circuit. A capacitance of the first connection line is compensated by electrically connecting the compensation portion to the first connection line, and thus a difference between the capacitance of the first connection line and a capacitance of the second connection line is reduced.

Abstract: Embodiments of the disclosure provide a comb drive, a comb drive system, and a method of operating the comb drive to rotate bi-directionally in a MEMS environment. An exemplary comb drive system may include a comb drive, at least one power source, and a controller. The comb drive may include a stator comb having a first electrically conductive layer spaced apart from a second electrically conductive layer. The comb drive may also include a rotor comb having a first electrically conductive layer spaced apart from a second electrically conductive layer. The controller may be configured to apply first and second voltage levels having opposite polarities to the first and second electrically conductive layers of the rotor comb, respectively. The controller may also be configured to apply an intermediate voltage level to one of the first or second electrically conductive layers of the stator comb.

Abstract: Embodiments of the disclosure include a receiver of an optical sensing system. The receiver may include a Hadamard mask configured to resonate during a scanning procedure performed by the optical sensing system. The Hadamard mask may include a frame beginning pattern corresponding to a start of a frame captured during the scanning procedure. The Hadamard mask may also include a coded pattern configured to provide sub-pixelization of the frame. The receiver may also include a photodetector array positioned on a first side of the Hadamard mask. The photodetector array may be configured to detect light that passes through the Hadamard mask during the scanning procedure to generate the frame.

Abstract: Embodiments of the disclosure provide an apparatus for emitting laser light and a system and method for detecting laser light returned from an object. The system includes a transmitter and a receiver. The transmitter includes one or more laser sources, at least one of the laser sources configured to provide a respective native laser beam having a wavelength above 1,100 nm. The transmitter also includes a wavelength converter configured to receive the native laser beams provided by the laser sources and convert the native laser beams into a converted laser beam having a wavelength below 1,100 nm. The transmitter further includes a scanner configured to emit the converted laser beam to the object in a first direction. The receiver is configured to detect a returned laser beam having a wavelength below 1,100 nm and returned from the object in a second direction.

Abstract: Provided are the crystalline solid meglumine salts of (R)-5-(4-chlorophenyl)-1-isopropyl-2-methyl-4-(3-(4-(4-((4-((1-(phenylthio)-4-(4-((phosphonooxy)methyl)piperidin-1-yl)butan-2-yl)amino)-3- ((trifluoromethyl)sulfonyl)phenyl)sulfonamido)phenyl) piperazin-1-yl)phenyl)-1H-pyrrole-3-carboxylic acid. Pharmaceutical compositions having one or more of the crystalline solid meglumine salt compounds and methods for administering the crystalline solid meglumine salt compounds to a subject are also described. Methods for preparing the crystalline solid meglumine salt compounds are also provided.

Abstract: Embodiments of the disclosure provide a system for analyzing noise data for light detection and ranging (LiDAR). The system includes a communication interface configured to sequentially receive noise data of the LiDAR in time windows, at least one storage device configured to store instructions, and at least one processor configured to execute the instructions to perform operations. Exemplary operations include determining an estimated noise value of a first time window using the noise data received in the first time window and determining an instant noise value of a second time window using the noise data received in the second time window. The second time window is immediately subsequent to the first time window. The operations also include determining an estimated noise value of the second time window by aggregating the estimated noise value of the first time window and the instant noise value of the second time window.

Abstract: A plurality of modules are arranged on the broadside of a ship, each module includes a housing, and the housing is provided with a composite gas supply assembly, a composite airbag assembly, an intelligent control assembly, and a mechanical transmission assembly; the intelligent control assembly is configured for predicting the impact energy and a collision angle of a collision object, sending an inflation instruction to the composite gas supply assembly, inflating a large airbag to a proper air pressure, inflating a small airbag to a rated air pressure, and sending an angle adjustment instruction to the mechanical transmission assembly to adjust the collision angle of the small airbag; the large airbag is matched with the small airbag to fully absorb a collision force vertically acting on the surface of a hull.

Abstract: A modular cabin transfer device suitable for large luxury cruise ships includes transfer vehicles and pull rods, where each of the transfer vehicles includes a clamping device, clamping assemblies, and a driving assembly, the upper and lower ends of the clamping device are respectively provided with the clamping assemblies, the clamping device makes the clamping assemblies at the upper and lower ends thereof move toward each other or move backwards, and the clamping assemblies at the upper and lower ends thereof clamp the cabin; the driving assembly is arranged at the lower end of the clamping device, and is configured to drive the transfer vehicle to move; adjacent transfer vehicles are connected by the pull rods.

Abstract: In one example, a method of fabricating a polygon assembly of a Light Detection and Ranging (LiDAR) module is provided. The method comprises: forming, on a backside surface of a first silicon-on-insulator (SOI) substrate, a multi-facet polygon of the polygon assembly; forming, on a frontside surface of the first SOI substrate, an axial portion of a support structure of the polygon assembly, the axial portion forming a stack with the polygon along a rotation axis; forming, on a frontside surface of a second SOI substrate, a plurality of radial portions of the support structure; forming, on a backside surface of the second SOI substrate, a cavity that encircles the plurality of radial portions; and bonding, based on a wafer bonding operation, the axial portion to the plurality of radial portions to form the polygon assembly.

Abstract: Embodiments of the disclosure provide a collimating scanner for an optical sensing system, a method for fabricating the collimating scanner, and a transmitter that includes the collimating scanner. An exemplary collimating scanner may include a scanning mirror configured to steer a light beam towards an object. The collimating scanner may also include a Fresnel zone plate profile patterned on the scanning mirror configured to collimate the light beam. The disclosed collimating scanner eliminates the use a separate collimating lens and thus improves the form factor of the optical sensing system.

Abstract: Embodiments of the disclosure provide an optical sensing system, a range estimation system for the optical sensing system, and a method for the optical sensing system. The exemplary optical sensing system includes a transmitter configured to emit a laser pulse towards an object. The optical sensing system further includes a range estimation system configured to estimate a range between the object and the optical sensing system. The range estimation system includes an analog to digital converter (ADC) configured to generate a plurality of pulse samples based on the laser pulse returned from the object. The returned laser pulse has a substantially triangular waveform including a rising edge and a falling edge. The range estimation system further includes a processor. The processor is configured to generate synthesized pulse samples on the substantially triangular waveform based on the pulse samples.

Abstract: A fertilizer grinding mechanism, a water and fertilizer mixing device, and an aqueous fertilizer solution preparation production line. A fertilizer receiving body, where an edge is provided with a arranged first channel; a connecting body, fixed to a bottom of the fertilizer receiving body and provided with a second channel with the first channel, a bottom of the second channel is communicated with a buffer bin, a pushing block is arranged in the buffer bin, and the pushing block is connected with a drive mechanism for driving the block to move in direction of the body; and an inner grinding block, where the block is fixed to a bottom of the connecting body, an outer side of the inner grinding block is with a vertically movable grinding roller, the pushing block can push fertilizer under vertical movement of the roller. The grinding mechanism has a heat dissipation effect.

Abstract: Embodiments of the disclosure provide for a LiDAR system. The LiDAR system may generate a first FOV that is large and has rough resolution and a second FOV that is smaller and has a finer resolution. For an area of importance, such as along the horizon where pedestrians, vehicles, or other objects may be located, the second FOV with the finer resolution may be used. Using fine resolution for the area of importance may achieve a higher-degree of accuracy/safety in terms of autonomous navigation decision-making than if coarse resolution is used. Because the use of fine resolution is limited to a relatively small area, a reasonably sized photodetector and laser power may still be used to generate a long distance, high-resolution point-cloud.