Abstract: A federated learning method includes a first network device that obtains local training data of at least two terminal devices; processes the local training data of the at least two terminal devices, to obtain training datasets; performs model training based on the training datasets, to obtain a model gradient; and sends the model gradient to a second network device. In this way, training data of the first network device is from the at least two terminal devices.

Abstract: Embodiments of this application provide a communication method, apparatus, and system. The method includes: A network device performs time calibration on a first time unit of a first frequency band and a second time unit of a second frequency band; the network device sends first information to a terminal device based on the first frequency band and the first time unit, where the first information includes synchronization information; and the network device sends second information to the terminal device based on the second frequency band and the second time unit, where the second information includes information other than the synchronization information. For an obtained time window of the second information, refer to a time window of the first information.

Abstract: This application discloses a communication method that includes: receiving first control information of a first bandwidth part (BWP) and second control information of a second BWP, and receiving a first data channel and a second data channel based on the first control information. The first BWP includes a first frequency domain resource and a second frequency domain resource. The second BWP includes a third frequency domain resource, and the third frequency domain resource and the second frequency domain resource overlap. The first data channel is located on the first frequency domain resource, and the second data channel is located on the third frequency domain resource. The foregoing method provides a manner of transmitting data on the first and second BWPs that partially overlap.

Abstract: An array substrate, a method for manufacturing an array substrate, a display panel and a display device are provided. The array substrate includes a substrate; an insulating layer arranged at a side of the substrate, where the insulating layer is provided with an opening, the insulating layer includes a first area and a second area, a thickness of the first area is less than a thickness of the second area, and the first area is arranged around at least a part of the opening; and a first conductive part arranged at a side of the insulating layer away from the substrate, where an orthographic projection of the first conductive part on the substrate and an orthographic projection of the first area on the substrate are at least partially overlapped.

Abstract: A method of generating a 3D video, a method of training a neural network model, an electronic device, and a storage medium, which relate to a field of image processing, and in particular to technical fields of computer vision, augmented/virtual reality and deep learning. The method includes: determining, based on an input speech feature, a principal component analysis (PCA) coefficient by using a first network, wherein the PCA coefficient is used to generate the 3D video; correcting the PCA coefficient by using a second network; generating a lip movement information based on the corrected PCA coefficient and a PCA parameter for a neural network model, wherein the neural network model includes the first network and the second network; and applying the lip movement information to a pre-constructed 3D basic avatar model to obtain a 3D video with a lip movement effect.

Abstract: The disclosure provides a method for generating a three-dimensional virtual image. The method includes: obtaining a face image to be processed and a three-dimensional reference model; obtaining a three-dimensional face model, face attribute information and face image information by inputting the face image to be processed into a trained neural network; obtaining a three-dimensional image model by performing a deformation process on the three-dimensional reference model based on the three-dimensional face model; and obtaining a target virtual image by adjusting the three-dimensional image model based on the face attribute information and the face image information.

Abstract: A method and apparatus for predicting a mouth-shape feature, and an electronic device are provided. A specific implementation of the method comprises: recognizing a phonetic posterior gram (PPG) of a phonetic feature; and performing a prediction on the PPG by using a neural network model, to predict a mouth-shape feature of the phonetic feature, the neural network model being obtained by training with training samples and an input thereof including a PPG and an output thereof including a mouth-shape feature, and the training samples including a PPG training sample and a mouth-shape feature training sample.

Abstract: A method, an apparatus, a device and a computer-readable storage medium for recognizing aerial handwriting are provided. The method may include detecting a palm region of a user in a two-dimensional grayscale image; detecting a fingertip in the two-dimensional gray-scale image based on the palm area; determining a spatial trajectory of the fingertip based on a set of two-dimensional gray-scale images, the set of two-dimensional gray-scale images including the two-dimensional gray-scale image; and recognizing handwritten content of the user based on the spatial trajectory. A two-dimensional gray-scale image is used to recognize and track the spatial trajectory of the fingertip, which may speed up aerial handwriting recognition, and has low processing performance requirements for the device, while also ensuring high accuracy.

Abstract: A camera and a display panel are disposed on a same side of a terminal. The camera includes a lens barrel and a plurality of lenses disposed in the lens barrel in a stacked manner in a direction of an optical axis of the camera. The lens barrel includes a first segment housing a first lens, wherein the first lens includes a light incident surface, a light exiting surface, and a side surface connected between the light incident surface and the light exiting surface. The side surface includes a first surface and a second surface connected to the first surface, wherein a perpendicular distance between the first surface and the optical axis is less than a perpendicular distance between the second surface and the optical axis. A part of the first segment is disposed adjacent to the display panel to reduce a space occupied by the camera.

Abstract: The disclosure provides a method for generating a three-dimensional virtual image. The method includes: obtaining a face image to be processed and a three-dimensional reference model; obtaining a three-dimensional face model, face attribute information and face image information by inputting the face image to be processed into a trained neural network; obtaining a three-dimensional image model by performing a deformation process on the three-dimensional reference model based on the three-dimensional face model; and obtaining a target virtual image by adjusting the three-dimensional image model based on the face attribute information and the face image information.

Abstract: A camera and a display panel are disposed on a same side of a terminal. The camera includes a lens barrel and a plurality of lenses disposed in the lens barrel in a stacked manner in a direction of an optical axis of the camera. The lens barrel includes a first segment housing a first lens, wherein the first lens includes a light incident surface, a light exiting surface, and a side surface connected between the light incident surface and the light exiting surface. The side surface includes a first surface and a second surface connected to the first surface, wherein a perpendicular distance between the first surface and the optical axis is less than a perpendicular distance between the second surface and the optical axis. A part of the first segment is disposed adjacent to the display panel to reduce a space occupied by the camera.

Abstract: A method and apparatus for predicting a mouth-shape feature, and an electronic device are provided. A specific implementation of the method comprises: recognizing a phonetic posterior gram (PPG) of a phonetic feature; and performing a prediction on the PPG by using a neural network model, to predict a mouth-shape feature of the phonetic feature, the neural network model being obtained by training with training samples and an input thereof including a PPG and an output thereof including a mouth-shape feature, and the training samples including a PPG training sample and a mouth-shape feature training sample.

Abstract: A method, an apparatus, a device and a computer-readable storage medium for recognizing aerial handwriting are provided. The method may include detecting a palm region of a user in a two-dimensional grayscale image; detecting a fingertip in the two-dimensional gray-scale image based on the palm area; determining a spatial trajectory of the fingertip based on a set of two-dimensional gray-scale images, the set of two-dimensional gray-scale images including the two-dimensional gray-scale image; and recognizing handwritten content of the user based on the spatial trajectory. A two-dimensional gray-scale image is used to recognize and track the spatial trajectory of the fingertip, which may speed up aerial handwriting recognition, and has low processing performance requirements for the device, while also ensuring high accuracy.

Abstract: Systems and methods described herein provide for a process for managing continuity of experience between operation of a hybrid computer system in the connected state and the slate component independently in the disconnected state. Embodiments further provide for the continuity of experience for devices operating under multiple processors or multiple platforms. According to embodiments, one or more states and conditions of the connected hybrid computer system may be synchronized with the slate component when the slate component is disconnected from the hybrid computer system. Alternatively, embodiments provide for the synchronization of one or more states and conditions of the independent slate component with the hybrid computer system responsive to connecting the slate component to the hybrid computer system. Non-limiting examples of states and conditions according to embodiments are web pages, applications, documents, lists of recently opened files and web pages, and web browser active tabs.

Abstract: A method of determining the electroosmotic transport coefficient of a proton exchange membrane, the method including creating a stream of hydrated hydrogen on either side of the membrane which is permanently controlled so that the relative humidity is almost identical on each side of the membrane at any point, thereby making it possible to minimize any back diffusion into the membrane. Furthermore, the method includes estimating the back diffusion flux into the membrane from the rate of return to equilibrium of the relative humidity starting from the moment when the current is cut off.

Abstract: A method of determining the electroosmotic transport coefficient of a proton exchange membrane, the method including creating a stream of hydrated hydrogen on either side of the membrane which is permanently controlled so that the relative humidity is almost identical on each side of the membrane at any point, thereby making it possible to minimize any back diffusion into the membrane. Furthermore, the method includes estimating the back diffusion flux into the membrane from the rate of return to equilibrium of the relative humidity starting from the moment when the current is cut off.

Abstract: Systems and methods described herein provide for a process for managing continuity of experience between operation of a hybrid computer system in the connected state and the slate component independently in the disconnected state. Embodiments further provide for the continuity of experience for devices operating under multiple processors or multiple platforms. According to embodiments, one or more states and conditions of the connected hybrid computer system may be synchronized with the slate component when the slate component is disconnected from the hybrid computer system. Alternatively, embodiments provide for the synchronization of one or more states and conditions of the independent slate component with the hybrid computer system responsive to connecting the slate component to the hybrid computer system. Non-limiting examples of states and conditions according to embodiments are web pages, applications, documents, lists of recently opened files and web pages, and web browser active tabs.

Abstract: Methods, and systems for migrating personality of a computing environment from a source machine platform to a target machine platform through a CIM-based system management infrastructure. A system includes: database wherein migration rules of CIM objects between a plurality of platforms are recorded; and migration tool for implementing migration according to the migration rules in the database, which includes: migration rule extractor, CIM object extractor, migration task producer and migration task executor. The extractor extracts CIM objects related to personality of the computing environment from source machine platform according to migration rule extracted from the database by rule extractor, and extracts from the target machine platform objects corresponding to the extracted CIM objects of the source machine platform. The migration task producer generates the migration tasks according to extracted migration rules.

Abstract: Computer systems in which at least two operating systems coexist independently from each other, and that swap between the operating systems. Provides methods to have at least two operating systems coexist independently of each other and swap the operating systems. A computer system includes a memory allocation component to be called by a power-on self test (POST) routine in order to allocate RAM for the operating systems when the computer system starts up, and backup and restore component to be called by a suspension routine in order to backup and restore the data specific to the operating systems, wherein each of the operating systems can transfer itself to the memory area belonging to itself, turn an allocated memory area into a physical RAM, and call the suspension routine to swap to other operating systems by setting swap parameters. The computer system can swap between operating systems easily and conveniently.