利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

1117 条记录 (耗时 0.031 秒)

    HETEROGENEOUS GRAPH NEURAL NETWORK USING OFFSET TEMPORAL LEARNING FOR SEARCH PERSONALIZATION
    1.
    发明公开
    HETEROGENEOUS GRAPH NEURAL NETWORK USING OFFSET TEMPORAL LEARNING FOR SEARCH PERSONALIZATION 审中-公开

    公开(公告)号：US20240346309A1

    公开(公告)日：2024-10-17

    申请号：US18582249

    申请日：2024-02-20

    申请人： Roku, Inc.

    发明人： Abhishek Majumdar ,  Kapil Kumar ,  Nitish Aggarwal ,  Srimaruti Manoj Nimmagadda

    IPC分类号： G06N3/08 ,  G06N3/042

    CPC分类号： G06N3/08 ,  G06N3/042

    摘要： Disclosed herein are system, apparatus, article of manufacture, method and/or computer program product embodiments, and/or combinations and sub-combinations thereof, for training a heterogenous graph neural network (GNN) to generate user embeddings corresponding to users and item embeddings corresponding to items. An example embodiment generates a first user interaction graph for a first time window and a second user interaction graph for a second time window, wherein each graph represents users and items as nodes and user-item interactions within the respective time window as edges, samples user-item node pairs from the second user interaction graph, and trains the heterogeneous GNN based on user-item node pairs from the first user interaction graph that correspond to the sampled user-item node pairs from the second user interaction graph. User and item embeddings generated by the trained GNN may be used to determine a relevancy of a given item with respect to a given user.

    Environmentally aware remote control
    2.
    发明授权
    Environmentally aware remote control 有权

    公开(公告)号：US12117513B2

    公开(公告)日：2024-10-15

    申请号：US17332353

    申请日：2021-05-27

    申请人： Roku, Inc.

    发明人： Eric Casten Riepling

    IPC分类号： G01S11/14

    CPC分类号： G01S11/14

    摘要： Disclosed herein are system, apparatus, article of manufacture, computer-implemented method, and/or computer-program product (computer-readable non-transitory storage medium) embodiments, and/or combinations and sub-combinations thereof, for an environmentally aware remote control. An example embodiment operates by at least one processor detecting a first signal and a second signal and receiving information corresponding to at least a physical position where a user is located relative to at least the first physical source. The at least one processor may further operate by calculating a time difference between a given sample of the first signal and a corresponding sample of a second signal at the physical position where the user is located, and adjusting a relative timing of the first signal with respect to the second signal. In some further embodiments, the at least one processor may adjust other attributes of either signal besides timing, e.g., audio characteristics such as volume, pitch, other filtering, etc.

    VOICE CONTROL DEVICE WITH PUSH-TO-TALK (PTT) AND MUTE CONTROLS
    6.
    发明公开
    VOICE CONTROL DEVICE WITH PUSH-TO-TALK (PTT) AND MUTE CONTROLS 审中-公开

    公开(公告)号：US20240312458A1

    公开(公告)日：2024-09-19

    申请号：US18670478

    申请日：2024-05-21

    申请人： Roku, Inc.

    发明人： Dustin VERHOEVE ,  Ralph BIRT

    IPC分类号： G10L15/22 ,  H04N21/422

    CPC分类号： G10L15/22 ,  H04N21/42203 ,  H04N21/42204 ,  G10L2015/223

    摘要： Disclosed herein are system, apparatus, article of manufacture, method and/or computer program product embodiments, and/or combinations and sub-combinations thereof, for a voice control device including a microphone, a mute control, and a push-to-talk (PTT) control. An example embodiment operates by: entering a mute state from an always-listening state when the device receives a mute control signal; entering a PTT state from the mute state when the device is in the mute state and receives a first PTT control signal; activating the microphone when the device is in the PTT state; and entering the mute state from the PTT state when the device is in the PTT state and receives a second PTT control signal.

    Modifying graphics rendering by transcoding a serialized command stream
    9.
    发明授权
    Modifying graphics rendering by transcoding a serialized command stream 有权

    公开(公告)号：US12079899B2

    公开(公告)日：2024-09-03

    申请号：US18144793

    申请日：2023-05-08

    申请人： ROKU, INC.

    发明人： Matthew James Sottek

    IPC分类号： G06T1/60 ,  G09G5/39

    CPC分类号： G06T1/60 ,  G09G5/39

    摘要： Disclosed herein are system, method, and computer program product embodiments for modifying graphics rendering by transcoding a serialized command stream. An embodiment operates by receiving a command configured to instruct an API to render a graphics element. The embodiment further operates by generating, based on the command, a transcoded command configured to instruct the API to render a modified graphics element by applying a set of modification factors to a portion of the command. Subsequently, the embodiment operates by transmitting the transcoded command to the API.

    Speaker-identification model for controlling operation of a media player
    10.
    发明授权
    Speaker-identification model for controlling operation of a media player 有权

    公开(公告)号：US12075119B2

    公开(公告)日：2024-08-27

    申请号：US18189701

    申请日：2023-03-24

    申请人： Roku, Inc.

    发明人： Frank Maker

    IPC分类号： H04N21/45 ,  H04N21/422 ,  H04N21/4415 ,  H04N21/472

    CPC分类号： H04N21/4532 ,  H04N21/42203 ,  H04N21/4415 ,  H04N21/47202

    摘要： In one aspect, an example method includes (i) obtaining, by a media player of a media presentation system, an audio signal, where the audio signal includes a voice command and is obtained using a microphone of the media presentation system; (ii) identifying, by the media player, which of multiple speakers of a household uttered the voice command using the audio signal and a speaker-identification model; (iii) performing, by the media player, an action corresponding to the voice command; and (iv) based on the identifying of the speaker using the audio signal and the speaker-identification model, selecting, by the media player, a user profile associated with the identified speaker within a streaming channel so as to bypass a profile selection screen of the streaming channel.