Abstract: A semantic-driven immersive telepresence system for bandwidth-efficient transmission of three-dimensional human representations. The system extracts compact semantic details from captured image data instead of transmitting complete raw 3D content. Color semantics are derived by estimating scene lighting characteristics using a lightweight model, combined with pre-captured base skin and clothing color. Motion semantics are extracted via keypoint detection of body joints, facial features, and hand positions. A regression-based method using keypoint-anchored vertices (KAVs) maps sparse keypoints to parametric human model parameters in real-time. The system transmits compressed semantic information over networks and reconstructs human representations at receiving devices. A multi-stage neural rendering pipeline with patch-based processing generates photo-realistic visual output.

Abstract: Disclosed in embodiments of the present application are a resource allocation method, a device, and a medium. The method comprises: a terminal device determines N NPRACH transmission resources corresponding to a sending period of an ephemeris-bearing SIB bearing according to said sending period as well as configuration information for an NPRACH transmission resource indicated in an SIB2, where N is a natural number greater than zero; the terminal device randomly selects one NPRACH transmission resource from among the N NPRACH transmission resources; and the terminal device performs random access on a network device by means of the selected NPRACH transmission resource.

Abstract: The technical solutions of this application provide a video playing method and a device, so that in a multi-view mode, during video playback, audio corresponding to an image in a plurality of views may be selected and played based on an operation performed on a play control. In response to a user input, a video playing interface is displayed on the electronic device. The video playing interface includes a first video, a second video, a first play control, and a second play control. In response to an operation performed on the second play control, the first video and the second video are played. Audio of a video corresponding to a location of the first play control is played. The video corresponding to the location of the first play control is the first video.

Abstract: An intrinsically safe battery assembly includes a battery pack including at least one battery and a protection circuit connected in series with the battery pack. The protection circuit includes the following elements connected in series: at least one diode configured to prevent the battery pack from being reversely charged, at least one current limiting resistor configured to provide short circuit protection and a resettable fuse configured to provide over current protection.

Abstract: Generating images of items with artificial intelligence (AI)-generated backgrounds is described. An example process includes capturing an image(s) of an item, obtaining first image data that identifies a real-world background in the image(s), and presenting a user interface for a user of an electronic device to indicate a descriptor(s). The process may further include receiving, via the user interface, an indication of the descriptor(s), generating prompt data representing a prompt(s) based at least in part on the descriptor(s), sending the first image data and the prompt data to a server computer(s), receiving, from the server computer(s), second image data output by a trained AI model(s), the second image data representing an AI-generated image(s) with an AI-generated background(s), and presenting, based at least in part on the second image data, a candidate image(s) with the AI-generated background(s) for selection by the user.

Abstract: Aspects of the subject disclosure may include, for example, a device that has a processing system including a processor; and a memory that stores executable instructions that, when executed by the processing system, facilitate performance of operations, including downsampling a full point cloud to obtain a downsampled point cloud, wherein the downsampling reduces a data size of the full point cloud; and using a machine-learning model to assign labels for segmentation and object identification to points in the downsampled point cloud, wherein the machine-learning model is trained on the full point cloud. Other embodiments are disclosed.

Abstract: The present invention belongs to the technical field of bridge protection. In order to solve the problem of how to protect bridge piers from flow-induced vibration (FIV), the present invention provides a method and system for monitoring the bridge piers based on digital twin and a FIV protection device, comprising: an annular steel board and multiple polyurethane boards, an inner wall of the annular steel board is fixedly connected to a structure to be protected, hydraulic devices and multiple energy-dissipation springs are arranged between an outer wall of the annular steel board and the polyurethane boards, and the polyurethane boards are spaced apart. The present invention may reduce or even eliminate the FIV caused by seawater or rivers to the structure to be protected, thereby slowing down the fatigue damage caused by FIV to the structure to be protected.

Abstract: A method includes receiving a record containing clinical information associated with a patient; and automatically processing the record using a multi-stage Artificial Intelligence (AI) engine to generate at least one suggested code for one or more portions of the clinical information; wherein the multi-stage AI engine comprises: a first module configured to identify content highlights within the one or more portions of the clinical information and associate the content highlights with candidate codes, respectively; and a second module configured to remove ones of the content highlights and/or ones of the candidate codes that have been identified and/or associated in error to generate corrected content highlights that are associated with the corrected candidate codes, respectively.

Abstract: A client device or an online system communicating with the device receives video data captured by a camera of the device, in which the video data depicts a field of view of a display area of the device. The device/system detects an object within the field of view based on the video data and applies one or more machine-learning algorithms to identify the object as an item available at a source. The device/system accesses item data for items available at the source and selects one or more supplemental items associated with the identified item based on item data for the identified item and each supplemental item. The device/system generates an augmented reality element including a listing of the supplemental item(s), as well as information or a selectable option associated with each supplemental item. The augmented reality element is then displayed in the display area of the device.

Abstract: An anti-rotation structure includes a first elongated member including an outer surface having a first extension extending therefrom and a second extension extending therefrom. A second member is coupled to the first extension and is coupled to the second extension. The second member has a rib disposed against the outer surface.

Abstract: A method includes receiving a record containing clinical information associated with a patient; processing the record using a first AI model that is trained based on medical term frequency throughout respective ones of a plurality of historical medical records; processing the record using a second AI model that is trained based on medical terms used in encounter portions of the respective ones of the plurality of historical medical records; generating a first recommendation for performing a coding review of the record based on processing the record using the first AI model; generating a second recommendation for performing the coding review of the record based on processing the record using the second AI model; and logically combining the first and second recommendation for performing the coding review to generate a final recommendation for performing the coding review.

Abstract: An echo filtering method, an electronic device, a computer-readable storage medium, and an echo filtering apparatus are disclosed. The electronic device includes M microphones and N speakers. M and N are integers greater than 1. The method includes: obtaining N speaker signals corresponding to the N speakers; obtaining M microphone signals corresponding to the M microphones; and performing at least direct sound filtering on the N speaker signals and the M microphone signals to obtain a target signal. By using this method, better echo filtering effect can be obtained.

Abstract: An indication method and an indication device are provided. The method includes obtaining satellite parameters of a current satellite cell and one or more neighbor satellite cells, wherein the current satellite cell is a satellite cell where a terminal is located, the one or more neighbor satellite cells are satellite cells neighboring the current satellite cell; determining a basic list and a detailed list according to the satellite parameters of the current satellite cell and the neighbor satellite cells, wherein the basic list includes basic information of the neighbor satellites, and the detailed list includes detailed information of the neighbor satellites; sending the basic list and/or the detailed list to the terminal.

Abstract: Embodiments of the present invention provide a stereo sound pickup method and apparatus, a terminal device, and a computer-readable storage medium. The terminal device obtains a plurality of pieces of target sound pickup data from sound pickup data of a plurality of microphones, obtains posture data and camera data of the terminal device, determines, from a plurality of prestored beam parameter groups based on the posture data and the camera data, a target beam parameter group corresponding to the plurality of pieces of target sound pickup data, and forms a stereo beam based on the target beam parameter group and the plurality of pieces of target sound pickup data.

Abstract: Aspects of the subject disclosure may include, for example, a device that has a processing system including a processor; and a memory that stores executable instructions that, when executed by the processing system, facilitate performance of operations, including downsampling a full point cloud to obtain a downsampled point cloud, wherein the downsampling reduces a data size of the full point cloud; and using a machine-learning model to assign labels for segmentation and object identification to points in the downsampled point cloud, wherein the machine-learning model is trained on the full point cloud. Other embodiments are disclosed.

Abstract: Embodiments of this application provide an audio processing method and a device. In a multi-channel video recording mode, a plurality of channels of video images and a plurality of channels of audio can be recorded simultaneously, and different audio can be played during video playback. A specific solution is as follows: After detecting an operation of opening a camera by a user, an electronic device displays a shooting preview interface, and then enters a multi-channel video recording mode. After detecting a shooting operation of the user, the electronic device displays a shooting interface, where the shooting interface includes a plurality of channels of video images. Then, the electronic device records the plurality of channels of video images, and records audio corresponding to each of the plurality of channels of video images based on a shooting angle of view corresponding to each channel of video image.

Abstract: An integrated multidirectional loading model test device for offshore wind turbines, including a water flume, a wave-making mechanism at first end of water flume, a top hood at top of water flume to form air duct, a first end of the top hood as air inlet end and a fan at second end of top hood, a seabed soil holding sink at a set position at bottom of water flume, a rotating disk at bottom of interior of seabed soil holding sink, a model barrel for placing seabed soil and the offshore wind turbine model is fixed on the rotating disk, and the rotating disk is connected to a driving mechanism to drive model barrel to rotate, so that a relative motion between the model barrel and the fixed water flume can be generated to achieve the wind and wave loads in different directions on the offshore wind turbine model.

Abstract: The invention provides a sensor, which includes a sensing surface; a housing including a recess, the sensing surface is located at the bottom of the recess, and the recess includes at least one side wall and an opening circumferentially surrounding the sensing surface. A compressed fluid nozzle is provided in the at least one side wall to nozzle compressed fluid from the compressed fluid nozzle to force the substance on the sensing surface to leave from the opening. This design makes it possible to refresh in situ with good efficiency without involving additional mechanical parts, thus reducing the risk of sensing surface damage, such as scratch wear, in the presence of wear particles.

Abstract: A method for handover of a satellite base station, a terminal, the satellite base station, and a storage medium are provided. The terminal determines a pitch angle when accessing the satellite base station; and if the pitch angle is determined within a preset range, the terminal performs inter-satellite handover; or the terminal sends the pitch angle to the accessed satellite base station, so that the satellite base station notifies the terminal to perform the inter-satellite handover when determining that the pitch angle is within the preset range.

Abstract: A repair device for burnout of a stator iron core in a giant hydraulic turbine generator, and a repair method thereof are provided. The repair device includes toothed plates, ventilation slots, binding holes, ventilation holes, and binding cables. Each of the toothed plates is matched with a burnout area in the middle of the iron core, such that two ends of the toothed plate are in contact with plate assemblies. The binding cables rotatably pass through the binding holes and are pulled tightly to bind the toothed plates and coil bars into an integral structure. The present disclosure solves the problem of influence on the power generation benefit during the flood season due to long re-lamination cycle after burnout of the middle of the stator iron core of the hydraulic turbine generator.