Abstract: A voice coil motor held immobile by magnetic attraction when not in use for photography purposes includes a base, a magnet fixed thereto, a carrier, a magnetic member, and a coil. The magnetic member is fixed to the carrier and is attracted by a permanent magnetic field (first MF) of the magnet on the base. The coil is fixed to the carrier and the magnetic member. When the voice coil motor is powered on, the coil generates a second magnetic field (second MF) opposing the first MF. The second MF reduces or eliminates the attractive force of the first MF of the magnet, to separate the carrier and its fittings and allow them to work freely. A lens module and an electronic device are further provided.

Abstract: A patch antenna comprises a multilayer printed circuit board that includes a calibration network, an array of patch radiators and a feed network. In some embodiments, the multilayer printed circuit board includes a plurality of dielectric substrates, wherein the array of patch radiators is provided on a dielectric substrate different from the dielectric substrate on which the calibration network is provided, and the dielectric substrate provided with the array of patch radiators is provided above the dielectric substrate provided with the calibration network.

Abstract: A dialog processing apparatus obtains dialog data, performs structured processing on the obtained dialog data, and outputs obtained structured data. The dialog data includes text data corresponding to a dialog performed between at least two dialog participants, key information in the dialog data is logically ordered in the structured data, and each piece of key information indicates a dialog topic of the dialog performed between the at least two dialog participants.

Abstract: In an embodiment, an amplifier circuit includes a second stage that includes a first switch circuit including first and second terminals, a plurality of resistive elements coupled between the first and second terminals of the first switch circuit, and a plurality of switches configured to control an equivalent resistance between the first and second terminals of the first switch circuit. During play mode, the second stage has a gain between the input of the second stage and the output of the second stage of a first value. During a transition from mute mode to play mode, the amplifier circuit is configured to progressively increase the gain of the second stage from a second value to the first value. During a transition from play mode to mute mode, the amplifier circuit is configured to progressively decrease the gain of the second stage from the first value to the second value.

Abstract: A method for dynamically calibrating a light detector of a light detection and ranging (LiDAR) system is disclosed. The method comprises obtaining an indication for use. The method further comprises determining, based on the indication, whether to perform the calibration of the light detector operating with a first bias voltage. The method further comprises, in accordance with a determination to perform the calibration, initiating a multiple-point calibration of the light detector across a bias voltage scanning range, wherein the multiple-point calibration comprises determining a second bias voltage corresponding to a current temperature in an operating environment of the light detector. The method further comprises determining, based on the multiple-point calibration, whether to update the first bias voltage based on the second bias voltage.

Abstract: Embodiments utilize trained query performance machine learning (QP-ML) models to predict an optimal compute node cluster size for a given in-memory workload. The QP-ML models include models that predict query task runtimes at various compute node cardinalities, and models that predict network communication time between nodes of the cluster. Embodiments also utilize an analytical model to predict overlap between predicted task runtimes and predicted network communication times. Based on this data, an optimal cluster size is selected for the workload. Embodiments further utilize trained data capacity machine learning (DC-ML) models to predict a minimum number of compute nodes needed to run a workload. The DC-ML models include models that predict the size of the workload dataset in a target data encoding, models that predict the amount of memory needed to run the queries in the workload, and models that predict the memory needed to accommodate changes to the dataset.

Abstract: Light detection and ranging (LiDAR) systems use light pulses to create an image or point cloud of an environment. This LiDAR system and method having data encoding and compression improves the efficiency and communication reliability of point cloud data using data encoding and compression. After receiving return light pulse reflected by an object in the FOV, the system converts the detected optical signal data into raw data for the purpose of generating trigger data, encoder data, and time synchronization data from the raw data. The system further configures output data in a compressed format using the least amount of bits to carry the same amount of information defining point cloud data describing an external environment and the computational load is reduced. The compressed format comprises a data set for one baseline channel and differential channel data for one or more channels based upon the baseline channel.

Abstract: Embodiments implement a prediction-driven, rather than a trial-driven, approach to automatic data placement recommendations for partitioning data across multiple nodes in a database system. The system is configured to extract workload-specific features of a database workload running at a database system and dataset-specific features of a database running on the database system. The workload-specific features characterize utilization of the database workload. The dataset-specific features characterize how data is organized within the database. The system identifies a plurality of candidate keys for determining how to partition data stored in the database across nodes. Based at least in part on the workload-specific features, the dataset specific features, and the plurality of candidate keys, a set of candidate key combinations for partitioning data is generated.

Abstract: A steam utilization system for producing a concrete pipe pile includes a steam conveying device, a steam curing pool, a plurality of reaction vessels, a steam generating device, a first steam distributor, a first pressure booster, a second pressure booster and a blowdown flash tank. A high-pressure steam inlet of the first steam distributor is communicated with the steam generating device, and a first steam inlet of the first pressure booster is communicated with a first high-pressure steam outlet of the first steam distributor. A second steam inlet of the first pressure booster is communicated with the steam conveying device, the steam outlet of the first booster is communicated with the steam inlet of each reaction vessel, a first steam inlet of a second pressure booster is communicated with the first steam distributor, a second steam inlet of the second pressure booster is communicated with the blowdown flash tank.

Abstract: The present disclosure relates to the technical field of cutting machining, and discloses a cross-scale structure feature surface machining method based on a multi-component collaborative vibration. A vibration in a z-axis direction is applied to a servo movement mechanism to realize the cutting of a micron-scale structure and the adjustment of the cutting depth; and the vibration in the z-axis direction is applied to a three-axis movement platform to realize the cutting of a millimeter-scale structure and the adjustment of the cutting depth. A required cross-scale structure feature surface can be machined and formed at one time through a collaborative vibration among a vibrating tool, a servo movement mechanism, and/or a three-axis movement platform according to the structure type contained in the required cross-scale structure, which can simplify a process flow and improve the machining efficiency, and has high economic efficiency.

Abstract: A light detection and ranging (LiDAR) system for scanning and reconfiguring regions-of-interest (ROIs) is provided. The system comprises a LiDAR scanner configured to scan a current set of ROIs within a field-of-view (FOV), and a LiDAR perception sub-system coupled to the LiDAR scanner. The LIDAR perception sub-system includes instructions for: obtaining sensor data provided at least by the LiDAR scanner; deriving one or more current perceptions based on the sensor data; obtaining one or more predefined perception policies; determining one or more policy-based ROI candidates; determining whether an ROI reconfiguration request is provided, based on a vehicle perception decision; determining one or more request-based ROI candidates based on the ROI reconfiguration request; and determining a next set of ROIs for the LiDAR scanner to scan based on the current set of ROIs, and one or both of the one or more policy-based ROI candidates and the one or more request-based ROI candidates.