FEDERATED LEARNING METHOD AND APPARATUS
    1.
    发明公开

    公开(公告)号:US20240054354A1

    公开(公告)日:2024-02-15

    申请号:US18493136

    申请日:2023-10-24

    CPC classification number: G06N3/098

    Abstract: This application provides a federated learning method and apparatus. The method includes: A first server receives a request message sent by at least one first client. The first server sends a training configuration parameter and a global model to the at least one first client. The first server receives first model update parameters separately fed back by the at least one first client. The first server aggregates the first model update parameters, to obtain first aggregation information in a current round of iteration. The first server obtains second aggregation information sent by the second server. The first server updates, based on the first aggregation information and the second aggregation information, the global model stored on the first server.

    APPLICATION COMPONENT DEPLOYMENT METHOD AND DEPLOYMENT NODE

    公开(公告)号:US20190213004A1

    公开(公告)日:2019-07-11

    申请号:US16278556

    申请日:2019-02-18

    Abstract: Embodiment of this application disclose an application component deployment method and a deployment node. In the method, a target deployment node receives a first deployment instruction sent by a management server, and determines a kinship node of the target deployment node according to the first deployment instruction, and a second application component that is in the multiple application components and that corresponds to the parent node, where the kinship node includes a parent node. Then, when detecting that the parent node has deployed the second application component, the target deployment node sends a second deployment instruction to the parent node. The target deployment node deploys a first application component according to the first deployment instruction.

    STEREO DECODING METHOD AND APPARATUS
    3.
    发明申请
    STEREO DECODING METHOD AND APPARATUS 有权
    立体声解码方法和装置

    公开(公告)号:US20160323687A1

    公开(公告)日:2016-11-03

    申请号:US15210644

    申请日:2016-07-14

    Abstract: A stereo decoding method and apparatus are disclosed. The method includes: restoring a monophonic signal from a received code stream through decoding; restoring an interchannel level difference, a group delay, and a group phase from the received code stream through decoding; and processing the monophonic signal according to the interchannel level difference, group delay, and group phase to obtain a first channel signal and a second channel signal. According to the stereo decoding method and apparatus provided in embodiments of the present invention, the first and second channel signals are obtained according to the monophonic signal, ILD, group delay, and group phase by referring to not only the ILD but also the group delay and group phase, thereby yielding favorable stereo sound field effect for the obtained first and second channel signals.

    Abstract translation: 公开了一种立体声解码方法和装置。 该方法包括:通过解码从接收到的码流中恢复单声道信号; 通过解码从接收到的代码流中恢复通道间级差,组延迟和组相位; 以及根据通道间电平差,组延迟和组相位处理单声道信号以获得第一通道信号和第二通道信号。 根据本发明的实施例中提供的立体声解码方法和装置,不仅参照ILD而且还参照组延迟,根据单声道信号,ILD,组延迟和组相位来获得第一和第二声道信号 和组相位,从而为获得的第一和第二信道信号产生有利的立体声场效应。

    SOUND IMAGE DIRECTION SENSE PROCESSING METHOD AND APPARATUS

    公开(公告)号:US20170223475A1

    公开(公告)日:2017-08-03

    申请号:US15487914

    申请日:2017-04-14

    Abstract: According to a sound image direction sense processing method and apparatus, a left-ear channel signal, a right-ear channel signal, and a centered channel signal that are of a sound source are obtained; whether a direction of the sound source is a front direction is determined according to the left-ear channel signal, the right-ear channel signal, and the centered channel signal; and when the direction of the sound source is the front direction, at least one type of the following processing: front direction enhancing processing or rear direction weakening processing is performed separately on the left-ear channel signal and the right-ear channel signal. Therefore, a difference between front direction sense and rear direction sense of a sound image may be enlarged, so that accuracy of determining a direction of a sound source may be improved.

    MODEL TRAINING METHOD, RELATED SYSTEM, AND STORAGE MEDIUM

    公开(公告)号:US20240119368A1

    公开(公告)日:2024-04-11

    申请号:US18540144

    申请日:2023-12-14

    CPC classification number: G06N20/00

    Abstract: Embodiments of this application provide a model training method. The system includes: A client is configured to train a first model based on unlabeled data, and is further configured to send a parameter of a first subnet in the first model to a server; the server is configured to train a second model based on the parameter of the first subnet and labeled data, to update a parameter of the second model, and the server is further configured to send an updated parameter of the first subnet and an updated parameter of a third subnet to the client; and the client is further configured to obtain a target model based on the parameter of the first subnet and the parameter of the third subnet that are from the server.

    OPTICAL/ELECTRICAL HYBRID CABLE AND OPTICAL COMMUNICATIONS SYSTEM

    公开(公告)号:US20220285050A1

    公开(公告)日:2022-09-08

    申请号:US17752292

    申请日:2022-05-24

    Abstract: Embodiments of this application provide an optical/electrical hybrid cable and an optical communications system, and relate to the field of optical communications technologies. The optical/electrical hybrid cable includes a linear conductor, an optical fiber, and an outer sheath. The outer sheath tightly wraps exteriors of the conductor and the optical fiber, and the conductor and the optical fiber are arranged side by side. According to the optical/electrical hybrid cable and the optical communications system provided in the embodiments of this application, on the premise that an optical fiber cable and a power cable that are separated can be formed, a structure is simple, a manufacturing process is simplified, and an area of a cross section of the optical/electrical hybrid cable is reduced.

    METHOD, DEVICE AND SYSTEM FOR DATA STREAM TRANSMISSION
    8.
    发明申请
    METHOD, DEVICE AND SYSTEM FOR DATA STREAM TRANSMISSION 有权
    数据流传输的方法,装置和系统

    公开(公告)号:US20140157014A1

    公开(公告)日:2014-06-05

    申请号:US14176712

    申请日:2014-02-10

    Abstract: The embodiments of the present disclosure provide a method, a device and a system for data stream transmission, which relate to the field of communication, and can reduce the energy consumption of the data stream transmission between devices. The method includes: receiving, by a resting device, a message indicating to enter into an energy-saving state; after receiving the message indicating to enter into the energy-saving state, sending, by the resting device, stored user information to an on-duty device so that the on-duty device receives and synchronizes user information; and, after the user information is synchronized, setting, by the resting device, the transmission state of the resting as standby and entering into a sleeping state. The embodiments of the present disclosure are used for data stream transmission.

    Abstract translation: 本公开的实施例提供了与通信领域有关的数据流传输的方法,设备和系统,并且可以减少设备之间的数据流传输的能量消耗。 该方法包括:由休息装置接收指示进入节能状态的消息; 在接收到指示进入节能状态的消息之后,由休息装置将存储的用户信息发送到占空比装置,使得值班装置接收并同步用户信息; 并且在用户信息同步之后,通过休息装置将休息的发送状态设置为待机状态并进入休眠状态。 本公开的实施例用于数据流传输。

Patent Agency Ranking