利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

5865 条记录 (耗时 0.025 秒)

    MACHINE LEARNING DATA AUGMENTATION USING DIFFUSION-BASED GENERATIVE MODELS
    1.
    发明公开
    MACHINE LEARNING DATA AUGMENTATION USING DIFFUSION-BASED GENERATIVE MODELS 审中-公开

    公开(公告)号：EP4462376A1

    公开(公告)日：2024-11-13

    申请号：EP24174841.7

    申请日：2024-05-08

    申请人： Siemens Healthineers AG

    发明人： GIBSON, Eli ,  MAILHE, Boris

    IPC分类号： G06V10/24 ,  G06V10/30 ,  G06V10/774 ,  G06V10/82 ,  G06N3/08 ,  G06T11/60 ,  G06T5/70

    摘要： Systems and methods for generating augmented images are provided. One or more input medical images are received. At least one of noise and one or more transformations are applied to the one or more input medical images to generate one or more noisy augmented images. The one or more noisy augmented images are denoised using a diffusion-based denoising system to generate one or more denoised augmented images. The applying and the denoising are repeated for one or more iterations using the one or more denoised augmented images as the one or more input medical images to generate one or more final augmented images. The one or more final augmented images are output.

    DATA PROCESSING METHOD AND RELATED DEVICE
    3.
    发明公开
    DATA PROCESSING METHOD AND RELATED DEVICE 审中-公开

    公开(公告)号：EP4462313A1

    公开(公告)日：2024-11-13

    申请号：EP23746144.7

    申请日：2023-01-17

    申请人： Huawei Technologies Co., Ltd.

    发明人： ZHOU, Pingyi ,  REN, Xiaozhe ,  WANG, Yasheng ,  HE, Bin ,  MENG, Xinfan ,  JIANG, Xin

    IPC分类号： G06N3/08

    摘要： This application relates to the artificial intelligence field, and discloses a data processing method. The method includes: processing target data through a target neural network to obtain a data processing result, where a target header of the target neural network is used to process, through a first transformation matrix, a first vector corresponding to first subdata, and process, through a second transformation matrix, a second vector corresponding to the first subdata, where the first vector corresponds to position information of the first subdata in the target data, and the second vector corresponds to semantic information of the first subdata. In this application, a matrix size of a transformation matrix corresponding to a position vector is set to be less than a size of a matrix corresponding to a semantic vector. To be specific, a size of the first transformation matrix is less than a size of the second transformation matrix. This can reduce a size of a transformation matrix used for calculating a correlation between position information, to reduce computing resource overheads of a model during inference or training.

    FEATURE ENCODING/DECODING METHOD AND DEVICE, AND RECORDING MEDIUM IN WHICH BITSTREAM IS STORED
    4.
    发明公开
    FEATURE ENCODING/DECODING METHOD AND DEVICE, AND RECORDING MEDIUM IN WHICH BITSTREAM IS STORED 审中-公开

    公开(公告)号：EP4462310A1

    公开(公告)日：2024-11-13

    申请号：EP22919069.9

    申请日：2022-12-22

    申请人： LG Electronics Inc.

    发明人： GWAK, Donggyu ,  LIM, Jaehyun ,  KIM, Chul Keun

    IPC分类号： G06N3/08 ,  G06N3/04 ,  H04N19/70 ,  H04N19/184

    摘要： Provided are a feature encoding/decoding method and device, and a computer-readable recording medium generated by the feature encoding method. The feature decoding method according to the present disclosure is performed by a feature decoding device, and comprises the steps of: acquiring, from a bit-stream, parameter information related to a plurality o layers; and recovering feature information about the plurality of layers on the basis of the parameter information, wherein the parameter information can include information about the number of feature channels and/or information about a quantization parameter (QP).

    SYSTEM AND METHOD FOR PERFORMING A CONDITIONAL HANDOVER
    8.
    发明公开
    SYSTEM AND METHOD FOR PERFORMING A CONDITIONAL HANDOVER 审中-公开

    公开(公告)号：EP4460085A1

    公开(公告)日：2024-11-06

    申请号：EP24173526.5

    申请日：2024-04-30

    申请人： Samsung Electronics Co., Ltd.

    发明人： SUNDARARAJU, Sathia Chandrane ,  DANDRA, Prasad Basavaraj ,  MADHURANTAKAM, Shrinath Ramamoorthy ,  VANAMA, Phanindhar

    IPC分类号： H04W36/00 ,  H04W36/36 ,  G06N3/08

    摘要： A method for performing a CHO includes receiving, via a network entity, a CHO reconfiguration associated with a plurality of target cells, identifying one or more target cells from among the plurality of target cells using the CHO reconfiguration, each of the one or more target cells having a higher signal strength than a serving cell associated with the UE, determining one or more handover trigger parameters associated with each of the one or more target cells using a network analysis-based Machine Learning model, the determining of the one or more handover trigger parameters being based on one or more network parameters, historical handover data, and one or more mobility parameters, selecting a first target cell from among the one or more target cells based on the one or more handover trigger parameters, and performing the CHO from the serving cell to the first target cell.

    AI EXERCISE GUIDE DEVICE AND METHOD
    9.
    发明公开
    AI EXERCISE GUIDE DEVICE AND METHOD 审中-公开

    公开(公告)号：EP4458437A1

    公开(公告)日：2024-11-06

    申请号：EP22916753.1

    申请日：2022-12-28

    申请人： DRAX Inc.

    发明人： YOO, Seon Kyung

    IPC分类号： A63B24/00 ,  A63B71/06 ,  G06N3/08 ,  G06N20/00 ,  G06V10/778

    摘要： As a preferred embodiment of the present disclosure, an AI exercise guide device includes a muscle strength estimation unit for estimating an estimated muscle strength value of a user's specific muscle strength based on user data, a detection unit for detecting a percentile value to which the estimated muscle strength value belongs among preset adult muscle strength percentile values, and a PMW estimation unit for estimating a PMW of exercise equipment the user intends to use based on the detected percentile value.