知搜-全球专利检索

  1. Login
  2. Sign Up

Search Results

29 records (took 0.022 seconds)

    Method and apparatus for training a convolutional neural network to detect defects
    21.
    发明授权
    Method and apparatus for training a convolutional neural network to detect defects 有权

    公开(公告)号：US11222234B2

    公开(公告)日：2022-01-11

    申请号：US16486025

    申请日：2019-04-02

    Applicant: BOE Technology Group Co., Ltd.

    Inventor： Tingting Wang

    IPC: G06K9/00 ,  G06K9/62 ,  G06K9/32 ,  G06T7/136 ,  G06T5/00 ,  G06N3/04 ,  G06N3/08 ,  G01N23/04 ,  G01N23/18 ,  G06T7/00

    Abstract: The present application discloses a method of training a convolutional neural network for defect inspection. The method includes collecting a training sample set including multiple solder joint images. A respective one of the multiple solder joint images includes at least one solder joint having one of different types of solder joint defects. The at least one solder joint is located substantially in a pre-defined region of interest (ROI) in a center of the image. The method further includes inputting the training sample set to a convolutional neural network to obtain target feature vectors respectively associated with the multiple solder joint images. Additionally, the method includes adjusting network parameters characterizing the convolutional neural network through a training loss function based on the target feature vectors and pre-labeled defect labels corresponding to different types of solder joint defects. The training loss function includes at least two different loss functions.

    Shift register, gate driving circuit and display panel for preventing channel short circuit
    23.
    发明授权
    Shift register, gate driving circuit and display panel for preventing channel short circuit 有权

    公开(公告)号：US12190782B2

    公开(公告)日：2025-01-07

    申请号：US18496888

    申请日：2023-10-29

    Applicant: Beijing BOE Display Technology Co., Ltd. ,  BOE TECHNOLOGY GROUP CO., LTD.

    Inventor： Tingting Wang ,  Qi Wang ,  Yan Yan

    IPC: G09G3/32 ,  G09G3/20 ,  G11C19/28 ,  G02F1/1362 ,  H01L27/12 ,  H01L29/417

    Abstract: The present disclosure provides a shift register, a gate driving circuit and a display panel. The shift register includes a transistor, which includes a gate electrode, a gate insulating layer, an active layer, a first electrode and a second electrode; the first and second electrode are of comb-shaped structures; the first electrode includes first and second comb tooth portions arranged at intervals, and a first comb handle portion connecting the first and second comb tooth portions; and comb tooth electrodes of the first comb tooth portions have a different length from those of the second comb tooth portions; the second electrode includes third and fourth comb tooth portions arranged at intervals, and a second comb handle portion connecting the third and fourth comb tooth portions; the first and third comb tooth portions form an inter-digital structure, the second and fourth comb tooth portions form an inter-digital structure.

    Display substrate, display panel and display apparatus
    26.
    发明授权
    Display substrate, display panel and display apparatus 有权

    公开(公告)号：US11586088B2

    公开(公告)日：2023-02-21

    申请号：US17599024

    申请日：2021-02-26

    Applicant: Beijing BOE Display Technology Co., Ltd. ,  BOE TECHNOLOGY GROUP CO., LTD.

    Inventor： Xue Wang ,  Xiao Wang ,  Yan Yan ,  Tingting Wang ,  Yaya Qi ,  Xiaoying Li ,  Zhiqiang Ma

    IPC: G02F1/1362 ,  G02F1/1368 ,  G02F1/1333 ,  G02F1/1343

    Abstract: The present disclosure provides a display substrate, a display panel and a display apparatus, belonging to the field of display technology. The display substrate includes a base, a plurality of common electrodes and a plurality of common electrode lines, the common electrodes are distributed on the base in an array, the common electrode lines extend along a row direction, and each common electrode line is connected to a corresponding row of common electrodes. The common electrode line is connected to the common electrode through a conductive connection portion, and the conductive connection portion includes conductive structures stacked on top of one another in a plurality of layers. The display substrate can reduce the resistance between the common electrode and the common electrode line, thereby reducing the voltage difference between the common electrodes in the display substrate and improving the uniformity of the common voltage therein.

    METHOD AND APPARATUS FOR TRAINING A CONVOLUTIONAL NEURAL NETWORK TO DETECT DEFECTS
    27.
    发明申请
    METHOD AND APPARATUS FOR TRAINING A CONVOLUTIONAL NEURAL NETWORK TO DETECT DEFECTS 有权

    公开(公告)号：US20210334587A1

    公开(公告)日：2021-10-28

    申请号：US16486025

    申请日：2019-04-02

    Applicant: BOE Technology Group Co., Ltd.

    Inventor： Tingting Wang

    IPC: G06K9/62 ,  G06T7/00 ,  G06K9/32 ,  G06T7/136 ,  G06T5/00 ,  G06N3/04 ,  G06N3/08 ,  G01N23/04 ,  G01N23/18

    Abstract: The present application discloses a method of training a convolutional neural network for defect inspection. The method includes collecting a training sample set including multiple solder joint images. A respective one of the multiple solder joint images includes at least one solder joint having one of different types of solder joint defects. The at least one solder joint is located substantially in a pre-defined region of interest (ROI) in a center of the image. The method further includes inputting the training sample set to a convolutional neural network to obtain target feature vectors respectively associated with the multiple solder joint images. Additionally, the method includes adjusting network parameters characterizing the convolutional neural network through a training loss function based on the target feature vectors and pre-labeled defect labels corresponding to different types of solder joint defects. The training loss function includes at least two different loss functions.

    Training method for tag identification network, tag identification apparatus/method and device
    28.
    发明授权
    Training method for tag identification network, tag identification apparatus/method and device 有权

    公开(公告)号：US11100369B2

    公开(公告)日：2021-08-24

    申请号：US16509774

    申请日：2019-07-12

    Applicant: BOE TECHNOLOGY GROUP CO., LTD.

    Inventor： Yue Li ,  Tingting Wang ,  Guangwei Huang

    IPC: G06K9/00 ,  G06K9/62 ,  G06K9/46 ,  G06N3/08

    Abstract: The present disclosure provides a training method for a tag identification network, a tag identification apparatus/method and device. The training method for a tag identification network includes: acquiring a first number of first images, each of the first images having its standard tag; adjusting parameters of the tag identification network by utilizing the first number of first images and their standard tags; selecting a second number of first images from the first number of first images as images to be converted, the second number being smaller than the first number; performing a style conversion process on each of images to be converted to generate a second image corresponding thereto, and serving a standard tag of the image to be converted as a standard tag of the second image; and adjusting the parameters of the tag identification network by utilizing the second number of the second images and their standard tags.

Patent Agency Ranking