Abstract: The present disclosure provides embodiments of semiconductor devices. In one embodiment, the semiconductor device includes a dielectric layer and a fin-shaped structure disposed over the dielectric layer. The fin-shaped structure includes a first p-type doped region, a second p-type doped region, and a third p-type doped region, and a first n-type doped region, a second n-type doped region, and a third n-type doped region interleaving the first p-type doped region, the second p-type doped region, and the third p-type doped region. The first p-type doped region, the third p-type doped region and the third n-type doped region are electrically coupled to a first potential. The second p-type doped region, the first n-type doped region and the second n-type doped region are electrically coupled to a second potential different from the first potential.

Abstract: This application discloses a channel estimation method and apparatus, a terminal, and a network-side device. The channel estimation method includes: receiving, by a terminal, a pilot signal sent by a network-side device, where resource blocks (RBs) occupied by the pilot signal in a first time domain transmission unit and RBs occupied by the pilot signal in a second time domain transmission unit are at least partially different; and performing channel estimation on a third time domain transmission unit based on the pilot signal in the first time domain transmission unit and the pilot signal in the second time domain transmission unit.

Abstract: A semiconductor device and a manufacturing method thereof are provided. The semiconductor device includes a heat transfer layer disposed over a substrate, a channel material layer, a gate structure and source and drain terminals. The channel material layer has a first surface and a second surface opposite to the first surface, and the channel material layer is disposed on the heat transfer layer with the first surface in contact with the heat transfer layer. The gate structure is disposed above the channel material layer. The source and drain terminals are in contact with the channel material layer and located at two opposite sides of the gate structure.

Abstract: A method of measuring physical properties includes a number of operations. Simulated mass flow rate data of training physical property parameter sets are generated by a computing unit based on the training physical property parameter sets and vessel shape information. Deep learning model is trained by a processor based on a training data set with an input feature vector of the simulated mass flow rate data and an output feature vector of the training physical property parameter sets. Measured fluid is received by a loader through an opening of a vessel of the vessel shape information. Weight accumulation data of weight of the loader is measured by a scale during a time period to obtain measured mass flow rate data. The measured mass flow rate data is inputted to the deep learning model to obtain measured physical property parameter set of the measured fluid.

Abstract: The present disclosure provides a semiconductor device. The semiconductor device includes a semiconductor substrate and an epitaxial stack disposed above the semiconductor substrate. The epitaxial stack includes first and second type epitaxial layers, the first and second type epitaxial layers having different material compositions. The first and second type epitaxial layers are alternatingly disposed in a vertical direction. The semiconductor device also includes a first doped region in the epitaxial stack and a second doped region in the epitaxial stack. The first doped region has a first dopant of a first conductivity type. The second doped region has a second dopant of a second conductivity type opposite the first conductivity type. The semiconductor device also includes first and second gate stacks disposed above the epitaxial stack. A portion of the first doped region and a portion of the second doped region are between the first and second gate stacks.

Abstract: Provided is the use of a silane composite emulsion as an anti-cracking enhancer in a cement-based material. In the present disclosure, active groups on an isobutyltriethoxysilane molecule react with hydroxyl groups on a surface of a mortar to form a layered hydrophobic structure, slowing down a water loss inside the mortar of a cement-based material. The dehydration of butyl acrylate forms a network structure, which fills pores of the cement-based material, inhibits shrinkage of the cement-based material, and reduces the cracking caused by the shrinkage. Tetraethoxysilane can undergo hydrolysis at a room temperature to generate nano-silica with a large number of hydroxyl groups on a surface; and the nano-silica can undergo secondary hydration with calcium hydroxide in the cement to form a secondary hydration product C—S—H gel, thereby filling most of voids to make a structure of the cement hydration product denser, to increase a strength of the mortar.

Abstract: Embodiments include a first set of fins having an emitter of a bipolar junction transistor (BJT) disposed over the first set of fins, a second set of fins having a base of the BJT disposed over the second set of fins, and a third set of fins having a collector of the BJT disposed over the third set of fins. A first gate structure is disposed over the first set of fins adjacent to the emitter. A second gate structure is disposed over the second set of fins adjacent to the base. A third gate structure is disposed over the third set of fins adjacent to the collector. The first gate structure, second gate structure, and third gate structure are physically and electrically separated.

Abstract: A miniature diaphragm-based fiber-optic tip FP pressure sensor, and fabrication method and application thereof. A miniature diaphragm-based fiber-optic tip FP pressure sensor includes an optical fiber, a hollow-core optical fiber, and a pressure sensing diaphragm, wherein the optical fiber and the hollow-core optical fiber have the same diameter, the two are spliced by arc welding; and the pressure sensing diaphragm is bonded to the endface of the hollow-core optical fiber by hydroxide catalysis bonding. The FP pressure sensor can not only realize the all-silica structure of a sensor, but also make the joint of each component free of organic polymer, and has extremely high long-term stability and thermal stability. Meanwhile, by a fabrication method of the miniature diaphragm-based fiber-optic tip FP pressure sensor, the application range and service life of the sensor are increased, and fabrication costs are reduced.

Abstract: A data query method is provided. The data query method includes receiving query description; processing, by a processor, the query description using one or more components; and determining query statement. Processing the query description includes obtaining query field from the query description; processing the query field using the one or more components; and determining a first sub-statement.

Abstract: A system generates a first variant of a first machine learning (ML) model, the first variant associated with an initial hyperparameter value. The system determines a prediction metric for the first variant of the first ML model, the prediction metric indicating an accuracy of the behavior prediction. The system generates an estimated simulation metric for the first variant of the first ML model by applying a second ML model to the prediction metric. The system identifies a first hyperparameter associated with a second variant of the first ML model, the second variant of the first ML model having a corresponding prediction metric and a corresponding estimated simulation metric that meet a first predetermined criteria, the second variant of the first ML model is used by an autonomous driving vehicle (ADV) for behavior prediction.

Abstract: A central air-conditioning system, comprises a plurality of air-conditioning assemblies (1) and a plurality of range hood assemblies (2); each air-conditioning assembly (1) comprises a compressor (11), a first heat exchanger (12) and a second heat exchanger (13) which are connected with each other through refrigerating medium pipes (14); all the range hood assemblies (2) are in communication with the public flue (6); a third heat exchanger (3) and the second heat exchanger (13) is connected through a secondary refrigerant channel (4); each end of each secondary refrigerant channel (4) respectively exchanges heat with the second heat exchanger (13) and the third heat exchanger (3).

Abstract: In one embodiment, a system determines a target lane for an autonomous driving vehicle (ADV) to change lanes from a current lane to the target lane. The system determines obstacles information for one or more obstacles surrounding the ADV from sensor data. The system determines vehicle information of the ADV including a speed of the ADV. The system applies a reinforcement learning (RL) model to the obstacles and vehicle information of the ADV to generate an action for the ADV, where the action includes an acceleration/deceleration value and a steering angle value. The system controls the ADV to perform the lane change from the current lane to the target lane by executing the action.

Abstract: Certain aspects of the present disclosure provide techniques and apparatus for beam selection using machine learning. A plurality of data samples corresponding to a plurality of data modalities is accessed. A plurality of features is generated by, for each respective data sample of the plurality of data samples, performing feature extraction based at least in part on a respective modality of the respective data sample. The plurality of features is fused using one or more attention-based models, and a wireless communication configuration is generated based on processing the fused plurality of features using a machine learning model.

Abstract: A long-distance lane change planning of an ADV is performed. A driving environment is perceived based on sensor data obtained from a plurality of sensors mounted on an ADV, including obtaining information of one or more obstacles on an adjacent lane. In response to a request to make a lane change from a current lane on which the ADV is driving to the adjacent lane, an S-V map is generated based on the information of the one or more obstacles. Each point on the S-V map represents a state of the ADV including a distance and a speed of the ADV. A trajectory of the ADV is generated using dynamic programming based on the S-V map. The ADV is controlled to drive autonomously according to the trajectory to make the lane change to the adjacent lane and avoid the one or more obstacles.

Abstract: Various embodiments of the teachings herein include a method for deploying power quality monitoring (PQM) devices. The method may include: determining a maximum number of PQM devices and historical power data of candidate deployment points, wherein the number of the candidate deployment points is greater than the maximum number of the PQM devices; clustering the historical power data of the candidate deployment points, wherein a target number of categories is determined on the basis of a silhouette coefficient of each candidate number of categories and the maximum number of the PQM devices; and determining PQM device deployment points based on the center of each category in the target number of categories.

Abstract: A bottom touching assisting device suitable for deep-sea submersibles and an implementation method thereof are provided. Four support columns are arranged within a mounting box body at a bottom of a deep-sea submersible, the support columns and the mounting box body are connected in a sliding manner through sliders, and the support column is sleeved with a threaded sleeve in a threaded connection manner. In conjunction with a drive component and a pressing mechanism, smooth vertical movement of the support column is achieved when the threaded sleeve rotates. This allows a bottom end of the support column to extend from the mounting box body.

Abstract: A computer-implemented method for image processing includes obtaining, based on a source image including a first biological object, an identity feature and attribute feature used to represent identity information and attribute information of the first biological object. Based on a target image including a second biological object, a context feature used to represent context information of the second biological object is obtained. Based on the identity feature, a fused image of the source image and the target image, the attribute feature, and the context feature is generated, where the fused image includes a fused biological object of a first biological object and the second biological object, identity information and attribute information of the fused biological object is identical to the identity information and the attribute information of the first biological object, and context information of the fused biological object is identical to the context information of the second biological object.

Abstract: An interface of a device is used to couple to another device and includes a set of data pins to support high speed data communication on an interconnect link between the devices based on an interconnect protocol. The interface further includes at least one auxiliary pin to support a particular signal defined by the interconnect protocol. The device is further configurated to generate hint data for use by the other device and send the hint data as a sideband signal to the other device over the auxiliary pin, where the sideband signal is distinct from signals defined for the auxiliary pin by the interconnect protocol.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training a machine learning model. A method includes: maintaining a plurality of training sessions; assigning, to each worker of one or more workers, a respective training session of the plurality of training sessions; repeatedly performing operations until meeting one or more termination criteria, the operations comprising: receiving an updated training session from a respective worker of the one or more workers, selecting a second training session, selecting, based on comparing the updated training session and the second training session using a fitness evaluation function, either the updated training session or the second training session as a parent training session, generating a child training session from the selected parent training session, and assigning the child training session to an available worker, and selecting a candidate model to be a trained model for the machine learning model.

Abstract: The invention discloses a double-station gantry combined machining system allowing automatic overturning of workpieces, which comprises a first mobile gantry machining unit, a second mobile gantry machining unit, an overturning platform, a first workpiece fixing table, a second workpiece fixing table and a guide rail, wherein the overturning platform is positioned between the first gantry machining unit and the second gantry machining unit. According to the invention, the system is highly integrated, continuous machining of a gantry machine tool is completed, the working efficiency is improved, and the machining cost of the system is reduced.