Abstract: The present disclosure provides a method for optimizing mask parameters, and the method includes: acquiring a test pattern, light source parameters, and initial mask parameters, the initial mask parameters including a mask thickness and an initial mask sidewall angle; generating multiple sets of candidate mask parameters according to the initial mask sidewall angle in the initial mask parameters; the multiple sets of candidate mask parameters including different mask sidewall angles and the same mask thickness; obtaining an imaging contrast of each set of candidate mask parameters based on the test pattern and the light source parameters; and selecting an optimal mask sidewall angle from the multiple sets of candidate mask parameters according to the imaging contrasts.

Abstract: In an IEEE 802.11 wireless system, a wireless STA is configured to signal a plurality of different spatial stream encoding values to a second STA device in a wireless personal area network in accordance with IEEE 802.11 protocol by generating a signaling physical layer protocol data unit (PPDU) which comprises a dedicated user information field which directly or indirectly identifies a plurality of different modulation and coding selection (MCS) values or different modulation orders used to encode different spatial streams from a plurality of spatial streams Nss used in a beamformed MIMO transmission of a data packet for each scheduled user of the signaling PPDU.

Abstract: Systems and methods for training a machine learning based segmentation network are provided. A set of medical images, each depicting an anatomical object, in a first modality is received. For each respective medical image of the set of medical images, a synthetic image, depicting the anatomical object, in a second modality is generated based on the respective medical image. One or more augmented images are generated based on the synthetic image. One or more segmentations of the anatomical object are performed from the one or more augmented images using a machine learning based reference network. An uncertainty associated with segmenting the anatomical object from the respective medical image is computed based on results of the one or more segmentations. It is determined whether the respective medical image is suitable for training a machine learning based segmentation network based on the uncertainty.

Abstract: Provided are a data apparatus and interaction method thereof, an electronic device, and a storage medium. The data interaction apparatus includes an interaction module, a switch module, a positioning module, a control module, and a drive module. The positioning module is configured to generate positioning information of an interaction object when the interaction object is located in a positioning detection region. The control module determines a corresponding target interaction unit according to the positioning information, and controls a target switch unit corresponding to the target interaction unit to communicate the target interaction unit with the drive module. The drive module drives the target interaction unit to perform data interaction with the interaction object, and sends data interaction information to the control module.

Abstract: A metal product includes a metal substrate, at least one first hole, at least one second hole, and at least one third hole. The first hole is formed in a surface of the metal substrate. The second hole is formed in at least one of a portion of the surface of the metal substrate without the first hole and an inner surface defining the first hole. The third hole is formed in at least one of a portion of the surface of the metal substrate without the first hole and without the second hole, a portion of the inner surface defining the first hole without the second hole, and an inner surface defining the second hole. The first, second, and third holes enhance a bonding strength between the metal product and a material product. The disclosure also provides a metal composite and a method for manufacturing the metal product.

Abstract: A thin film transistor includes a base, a first electrode, an active pattern, a gate insulating layer, a gate and a second electrode. The active pattern includes a first semiconductor pattern, a second semiconductor pattern and a third semiconductor pattern. A material of one of the first semiconductor pattern and the third semiconductor pattern includes a semiconductor material and N-type doped ions, and a material of another of the first semiconductor pattern and the third semiconductor pattern includes the semiconductor material and P-type doped ions. An orthogonal projection of the gate on the base is non-overlapping with an orthogonal projection of the active pattern on the base.

Abstract: Embodiments of the technology described herein describe a machine classifier capable of continually learning new classes through a continual few-shot learning approach. A natural language processing (NLP) machine classifier may initially be trained to identify a plurality of other classes through a conventional training process. In order to learn a new class, natural-language training data for a new class is generated. The training data for the new class may be few-shot training data. The training also uses synthetic training data that represents each of the plurality of other classes. The synthetic training data may be generated through a model inversion of the original classifier. The synthetic training data and the natural-language training data are used to retrain the NLP classifier to identify text in the plurality of other classes and the new class using.

Abstract: Methods, apparatuses and systems for using at least one sub-channel of a physical channel for uplink communication, the physical channel including a set of resources within first and second channel boundaries such that the physical channel includes a plurality of sub-channels, each sub-channel comprising a subset of the resources of the physical channel and having at least one sub-channel boundary which is not coincident with the first or the second channel boundaries, are provided. One representative apparatus includes a Station (STA) including: a processor, and a transmit/receive unit configured to: transmit, via an uplink, control information including a sub-channel SIG field indicating an identity of the STA, receive a trigger frame indicating one or more sub-channels available form among a plurality of sub-channels, determine which of the one or more sub-channels to acquire, and transmit data, via the uplink, using the one or more acquired sub-channels.

Abstract: Methods, systems, and devices for wireless communication are described. Neural networks may assist user equipments (UEs) and base stations in performing various operations related to wireless communications. For example, neural networks may be used to generate non-orthogonal cover codes for transmitting reference signals such as channel state information-reference signals (CSI-RSs). A base station may transmit, to a UE, a CSI-RS associated with a non-orthogonal cover code of a set of non-orthogonal cover codes. Using the CSI-RS, the UE may perform a channel estimation procedure that corresponds to the non-orthogonal cover code. Based on the channel estimation procedure, the UE may transmit a feedback message to the base station that indicates a channel quality parameter. Additionally, or alternatively, a UE may receive a CSI-RS, determine a precoding matrix using the CSI-RS and neural network, and transmit an indication of the pre-coding matrix to a base station.

Abstract: Embodiments of an apparatus and method for wireless communications are disclosed. In an embodiment, a wireless device includes a controller configured to construct an orthogonal matrix, and a wireless transceiver configured to conduct wireless communications based on the orthogonal matrix. The size of the orthogonal matrix is 12×12 or 16×16.

Abstract: An array substrate, a liquid crystal display panel and a display apparatus. The array substrate comprises: a substrate (10), a first insulating layer (20), a second insulating layer (30), a third insulating layer (40), a planarization layer (50), a first electrode layer (90A), a fourth insulating layer (70) and a second electrode layer (90B), the third insulating layer comprises a first interlayer insulating layer (40A), a second interlayer insulating layer (40B) and a third interlayer insulating layer (40C), which are sequentially stacked; the first interlayer insulating layer is located on the side of the second interlayer insulating layer close to the substrate (10), the third interlayer insulating layer is located on the side of the second interlayer insulating layer away from the substrate; the material of the first interlayer insulating layer and third interlayer insulating layer comprises silicon oxide, the material of the second interlayer insulating layer comprises silicon nitride.

Abstract: A display substrate includes a base substrate, a plurality of photosensitive transistor units, a plurality of photosensitive ESD protection units, and at least one common signal line. The base substrate includes a display region, a peripheral region located at a periphery of the display region, and a binding region located at a side of the display region. The plurality of photosensitive transistor units, the plurality of photosensitive ESD protection units and the at least one common signal line are located in the peripheral region. The plurality of photosensitive transistor units is connected with binding pins in the binding region through a plurality of signal lines. At least one photosensitive ESD protection unit is connected with, and located between, at least one signal line and the common signal line.

Abstract: The disclosure provides a flexible cryoablation needle device resistant to a low temperature and a high pressure, a threaded part is arranged on the outer wall of a liner pipe to enhance a connection strength between the liner pipe and a flexible pipe structure. Since an annular protrusion portion is arranged on the outer wall of the liner pipe, further leakage of a gas coming from a pressure relief process is prevented and the connection strength is enhanced as well. Air tightness and connection strength are further guaranteed by radial extrusion of extruding pipes. And an inner cavity in a cutter head is directly subjected to pressure relief through a cutter head vent, a pressure relief intermediate cavity, a liner vent, a flexible pipe vent and a pressure relief gap.

Abstract: The invention relates to a vehicle lighting system based on infrared light sources, the vehicle lighting system comprising: an infrared matrix lamp, which comprises a plurality of infrared light sources arranged in a matrix and which is configured to be mounted on a vehicle to provide an illumination range with respect to the vehicle, wherein each of the plurality of infrared light sources corresponds to an illumination region in the illumination range; and a matrix lamp controller configured to turn on, when a target region needs to be illuminated, an infrared light source in the infrared matrix lamp corresponding to the target region on the basis of relative position information between the target region and the vehicle, wherein the relative position information comprises an angular position of the target region relative to a vehicle reference. The invention further relates to a vehicle comprising the vehicle lighting system.

Abstract: A method for forming a glass sheet includes forming a glass ribbon. A robot arm is operated to move an end effector through a preprogrammed cycle. The cycle includes engaging a segment of the glass ribbon with the end effector, separating the engaged segment from the glass ribbon to generate a glass sheet, and moving the glass sheet away from the glass ribbon. The preprogrammed cycle designates predetermined positions of the end effector at predetermined points in time. While the robot arm is operating through the preprogrammed cycle, a parameter indicative of a force being exerted on the glass ribbon by the end effector is sensed. A position of the end effector is altered to differ from the predetermined position at the corresponding point in time when the sensed parameter deviates from a target value. An excessive force applied to the glass ribbon can be reduced in real-time.

Abstract: A light detection and ranging (LiDAR) system includes a rotatable polygon having a plurality of reflective sides including a first reflective side. The rotatable polygon configured to scan one or more first light signals in a first direction. The LiDAR system also includes a scanning optic configured to scan the one or more first light signals in a second direction different than the first direction. A first light source is configured to direct the one or more first light signals to one or more of the plurality of reflective sides of the rotatable polygon or the scanning optic. A first detector is configured to detect a first return light signal associated with a signal of the one or more first light signals. One or more optics are configured to focus the first return light signal on the first detector.

Abstract: A method includes obtaining a source training dataset that includes a plurality of source training images and obtaining a target training dataset that includes a plurality of target training images. For each source training image, the method includes translating, using the forward generator neural network G, the source training image to a respective translated target image according to current values of forward generator parameters. For each target training image, the method includes translating, using a backward generator neural network F, the target training image to a respective translated source image according to current values of backward generator parameters. The method also includes training the forward generator neural network G jointly with the backward generator neural network F by adjusting the current values of the forward generator parameters and the backward generator parameters to optimize an objective function.

Abstract: A pavement anti-skid performance evaluation method based on envelope feature is provided. A surface profile image of a to-be-tested piece is acquired as a blank background image. A motion trajectory image of a sliding piece on the to-be-tested piece is acquired. The blank background image is treated to obtain a surface texture profile curve of the to-be-tested piece. The motion trajectory image is differenced with the blank background image to obtain a difference image, which is treated to obtain a trajectory envelope curve of motion of the sliding piece on the to-be-tested piece. An arithmetic mean of the trajectory envelope curve and an arithmetic mean of the surface texture profile curve are calculated, and the anti-skid performance is evaluated based on a height difference ?h between above two arithmetic means. A pavement anti-skid performance evaluation device based on envelope feature is also provided.

Abstract: Certain aspects of the present disclosure provide techniques for sounding and precoding uplink transmissions using one or more machine learning models. An example method generally includes generating a sounding reference signal (SRS) using an SRS deep neural network (DNN); transmitting, to a network entity, the generated SRS in one or more resource elements (REs); receiving, from the network entity, information identifying a precoding matrix to use for uplink transmissions on a shared channel; precoding uplink transmissions based on the identified precoding matrix; and transmitting, to the network entity, the precoded uplink transmissions on the shared channel.

Abstract: A method for determining lower radius limit of movable throat of shale is provided, and it includes: performing a low-temperature nitrogen adsorption test on a target shale to obtain first pore radii; performing a high-pressure mercury injection test on the target shale to obtain second pore radii; performing a nuclear magnetic resonance test on the target shale to obtain third pore radii; obtaining a relationship diagram of distribution frequencies and pore radii according to three pore radii; distinguishing, according to the pore radii, relationship diagram data, and performing normalization processing to determining a relationship curve of normalized frequency data and the pore radii; and determining a lower radius limit of movable throat of shale according to relationship curve. A problem of describing characteristics of shale occurrence space with complex pore structures and strong heterogeneity is solved, the method is suitable for determining lower radius limit of movable throat of shale.