Abstract: Methods of training a machine learning model and using the trained model to assess ultrasound measurement data are disclosed. In one arrangement, training data comprises a plurality of classified frames of ultrasound measurement data. Each of at least a subset of the classified frames is classified as representing an imaging plane capable of providing information about a respective target anatomical feature. First and second samples of frames are selected. A machine learning model derives prototype feature vectors from the first sample and feature vectors from the second sample. A loss function depending on metrics representing distances between the feature vectors and the prototype feature vectors is optimized to train the machine learning model.

Abstract: A method for obtaining a higher-order ambisonics (HOA) coefficient includes obtaining location information of a virtual speaker on a preset spherical surface, where the preset spherical surface includes M circles of longitude and N circles of latitude, and obtaining, based on the location information and a preset reference trigonometric function table, a trigonometric function value corresponding to the location information, where the reference trigonometric function table includes an elevation trigonometric function table and/or an azimuth trigonometric function table, and obtaining an HOA coefficient for the virtual speaker based on the trigonometric function value corresponding to the location information.

Abstract: Techniques are disclosed to use a robotic arm to palletize or depalletize diverse items. In various embodiments, data associated with a plurality of items to be stacked on or in a destination location is received. A plan to stack the items on or in the destination location is generated based at least in part on the received data. The plan is implemented at least in part by controlling a robotic arm of the robot to pick up the items and stack them on or in the receptacle according to the plan, including by for each item: using one or more first order sensors to move the item to a first approximation of a destination position for that item at the destination location; and using one or more second order sensors to snug the item into a final position.

Abstract: This application provides a method and an apparatus for determining a virtual speaker set. The method for determining a virtual speaker set includes: determining a target virtual speaker from F preset virtual speakers based on a to-be-processed audio signal, where each of the F virtual speakers corresponds to S virtual speakers, F is a positive integer, and S is a positive integer greater than 1; and obtaining, from a preset virtual speaker distribution table, respective position information of S virtual speakers corresponding to the target virtual speaker, where the virtual speaker distribution table includes position information of K virtual speakers, the position information includes an elevation angle index and an azimuth angle index, K is a positive integer greater than 1, F?K, and F×S?K. This application can improve audio signal playback effect.

Abstract: Techniques are disclosed to perform robotic handling of soft products in non-rigid packaging. In various embodiments, sensor data associated with a workspace is received. An action to be performed in the workspace using one or more robotic elements is determined, the action including moving an end effector of one of the robotic elements relatively quickly to a location in proximity to an item to be grasped; actuating a grasping mechanism of the end effector to grasp the item using an amount of force and structures associated with minimized risk of damage to one or both of the item and its packaging; and using sensor data generated subsequent to the item being grasped to ensure the item has been grasped securely. Control communications are sent to the robotic element via the communication interface to cause robotic element to perform the action.

Abstract: A method of skin detection for an object includes: acquiring an image of skin covering an outer surface of an object to be detected; evaluating the image of the skin to determine a skin lesion region of the object to be detected, in which the skin lesion region is an image region with a skin lesion feature in the image of the skin; determining a skin lesion attribute of the object to be detected based on the skin lesion feature of the skin lesion region and an object type of the object to be detected, in which the skin lesion attribute is used for describing a skin lesion generated on the object to be detected; and matching lesion data recorded in a lesion database based on the skin lesion attribute to determine a pathological result of the object to be detected.

Abstract: Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a user equipment (UE) may receive, from a network node, a single rank reference signal. The UE may obtain, for multiple candidate beams, one or more single rank measurements from the single rank reference signal. The UE may identify, for each respective candidate beam included among the multiple candidate beams and based at least in part on the one or more single rank measurements obtained from the single rank reference signal for the respective candidate beam, a multiple rank spectral efficiency metric. The UE may identify, from among the multiple candidate beams and based at least in part on the multiple rank spectral efficiency metric associated with each respective candidate beam, a beam to use to communicate with the network node. Numerous other aspects are described.

Abstract: The present application discloses a warp execution method used for SPs of an SM of a GPU and an associated GPU.

Abstract: Methods, apparatus and systems for cell selection or reselection based on ephemeris and/or configuration information are disclosed. In one embodiment, a method performed by a wireless communication device operating in a non-terrestrial network (NTN), the method including: obtaining configuration information associated with a plurality of candidate cells of the NTN, each of the plurality of candidate cells associated with a respective one of a plurality of candidate satellites; and based on the configuration information, adjusting a ranking of the plurality of candidate cells for cell selection or reselection.

Abstract: A finger for a robotic gripper may include a flexible backbone, a plurality of jamming layers, and a membrane bag. The backbone may have a first side, a second side, a third side, and a fourth side. The backbone may include a flexible beam, and a plurality of branches attached to the flexible beam and spaced apart from one another. Each branch may include a first end surface extending along the first side, and a second end surface extending along the second side. The first end surfaces may collectively extend along a majority of the first side, and the second end surfaces may collectively extend along a majority of the second side. The jamming layers may be positioned along the third side or the fourth side. The membrane bag may be positioned over the jamming layers.

Abstract: Provided are fan-out package structure and fan-out packaging method. Structure includes packaging element, plastic package body, first dielectric layer, and second dielectric layer, packaging element has first pad and second pad, height of first pad is lower than that of second pad, plastic package body plastic-seals packaging element, and end surface of first pad is exposed from surface of plastic package body, one side of first pad away from packaging element has first wiring layer, and first wiring layer has first solder ball; first dielectric layer is provided on one side of first wiring layer away from packaging element, end surface of second pad is exposed from surface of first dielectric layer, one side of second pad away from packaging element has second wiring layer, and second wiring layer has second solder ball; and one side of second wiring layer away from first wiring layer has second dielectric layer.

Abstract: In one embodiment, the present disclosure is directed to a bathroom mirror lighting system comprising a mirror, a light source, a sensor, and a control circuit operably coupled to the light source and the sensor. The control circuit receives a sensor signal from the sensor indicative of at least one of a user's distance from the sensor, a user's position, a user's posture, a motion of the user towards or away from the sensor, or a task being performed by the user. The control circuit then transmits a control signal to the light source causing the light source to alter its light output based on the sensor signal from the sensor.

Abstract: Embodiments of the present disclosure relate to speech tendency classification.

Abstract: The present application discloses a GPU and a method of the same. The GPU includes: a plurality of streaming multiprocessor (SMs), each including: a plurality of streaming processors (SPs), each including a register, wherein each SP has a predetermined upper bound of warp number, and the register has a predetermined upper bound of register capacity; and a global dispatcher, including: a register occupancy status table, for recording the warp number and an occupancy status of the register of each SP of each SM; a TB (TB) dispatch module, for dispatching the TB to a first SM of the SMs according to a warp type classification table and the register occupancy status table; and a warp dispatch module, for dispatching a plurality of warps to the plurality of SPs of the first SM according to the warp type classification table and the register occupancy status table.

Abstract: A GPU and a GPU method are provided. The GPU includes stream multiprocessors, an available hardware resource table, resource comparator, stream scheduler, and global dispatcher. Each stream multiprocessor executes at least one thread block. The available hardware resource table records hardware resources available to the stream multiprocessors. The resource comparator selects from first-priority kernel codes in the kernel streams at least one first dispatchable kernel code whose required hardware resources are less than hardware resources available to the stream multiprocessors according to the available hardware resource table. The stream scheduler selects from the at least one first dispatchable kernel code a kernel code as a selected kernel code. The global dispatcher dispatches thread blocks in the selected kernel code to the stream multiprocessors for execution and updates the available hardware resource table according to usage conditions of hardware resources of the stream multiprocessors.

Abstract: A stream multiprocessor, a GPU, and related methods are provided. The stream multiprocessor executes thread blocks. Each thread block includes warps. The stream multiprocessor includes stream processors and a local dispatcher. Each stream processor executes one or more warps. The local dispatcher includes a warp state table, a warp resource detection unit and a warp launching unit. The warp state table records dispatching states and processing states of warps of the thread blocks. The warp resource detection unit selects all the first warps of a first thread block and at least one second warp of a second thread block according to hardware resources available to the stream multiprocessor and hardware resources required for thread blocks. The warp launching unit dispatches the first warps to idle stream processors and at least one second warp to at least one idle stream processor.

Abstract: Embodiments of the present disclosure relate to feature crossing for machine learning. According to example embodiments of the present disclosure, a method comprises determining a semantic correlation relationship between a plurality of feature categories, the semantic correlation relationship indicating respective degrees of semantic correlation between respective pairs of feature categories among the plurality of feature categories; obtaining at least two features classified in at least two of the plurality of feature categories for machine learning; and performing feature crossing on the at least two features based on the semantic correlation relationship.

Abstract: A method of manufacturing an interconnect structure includes forming an opening through a dielectric layer. The opening exposes a top surface of a first conductive feature. The method further includes forming a barrier layer on sidewalls of the opening, passivating the exposed top surface of the first conductive feature with a treatment process, forming a liner layer over the barrier layer, and filling the opening with a conductive material.

Abstract: A UE may receive at least two baseband signals at two or more antenna or beam ports of the UE over a measurement period, and determine an amount of motion of the UE based on the received at least two baseband signals. In one aspect, the amount of motion of the UE may be determined based on a variation of at least one phase delta of the at least one channel. In another aspect, the amount of motion of the UE may be determined based on a correlation between the at least two baseband signals.