公开(公告)号：US12148200B2

公开(公告)日：2024-11-19

申请号：US17419726

申请日：2020-12-29

Applicant: BOE Technology Group Co., Ltd.

Inventor： Yunhua Lu ,  Hanwen Liu ,  Pablo Navarrete Michelini ,  Lijie Zhang ,  Dan Zhu

IPC: G06V10/77 ,  G06N3/045 ,  G06N3/08 ,  G06T3/4046 ,  G06V10/80 ,  G06V10/82

Abstract: A method for processing an image includes acquiring an input image, performing down-sampling and feature extraction on the input image by an encoder network to obtain multiple feature maps, and performing up-sampling and feature extraction on the multiple feature maps by a decoder network to obtain a target segmentation image. Processing levels between the encoder network and the decoder network for outputting feature maps with the same resolution are connected with each other. The encoder network and the decoder network each includes one or more dense calculation blocks, and at least one convolution module in any dense computation block includes at least one group of asymmetric convolution kernels.

公开(公告)号：US11954822B2

公开(公告)日：2024-04-09

申请号：US17419350

申请日：2020-10-13

Applicant: BOE TECHNOLOGY GROUP CO., LTD.

Inventor： Pablo Navarrete Michelini ,  Wenbin Chen ,  Hanwen Liu ,  Dan Zhu

IPC: G06T3/4046 ,  G06N3/08 ,  G06T3/4053 ,  G06T5/50

CPC classification number: G06T3/4046 ,  G06N3/08 ,  G06T3/4053 ,  G06T5/50

Abstract: An image processing method, an image processing device, a training method of a neural network, an image processing method based on a combined neural network model, a constructing method of a combined neural network model, a neural network processor, and a storage medium are provided. The image processing method includes: obtaining, based on an input image, initial feature images of N stages with resolutions from high to low, where N is a positive integer and N>2, performing, based on initial feature images of second to N-th stages, cyclic scaling processing on an initial feature image of a first stage, to obtain an intermediate feature image; and preforming merging processing on the intermediate feature image to obtain an output image. The cyclic scaling processing includes hierarchically-nested scaling processing of N−1 stages, and scaling processing of each stage includes down-sampling processing, concatenating processing, up-sampling processing, and residual link addition processing.

公开(公告)号：US11587343B2

公开(公告)日：2023-02-21

申请号：US17044275

申请日：2020-04-20

Applicant: BOE TECHNOLOGY GROUP CO., LTD.

Inventor： Lijie Zhang ,  Guannan Chen ,  Hanwen Liu ,  Dan Zhu

IPC: G06V30/244 ,  G06F40/109 ,  G06V30/32 ,  G06V30/28

Abstract: A method and a system for converting a font of a Chinese character in an image, a computer device and a medium are disclosed. A specific implementation of the method includes: acquiring a stroke of a to-be-converted Chinese character in the image and spatial distribution information of the stroke; and generating a Chinese character in a target font that corresponds to the to-be-converted Chinese character in the image according to the stroke of the to-be-converted Chinese character, the spatial distribution information of the stroke and standard stroke information of the target font, to replace the to-be-converted Chinese character.

公开(公告)号：US11461639B2

公开(公告)日：2022-10-04

申请号：US16336995

申请日：2018-08-16

Applicant: BOE TECHNOLOGY GROUP CO., LTD.

Inventor： Hanwen Liu ,  Pablo Navarrete Michelini

IPC: G06V10/00 ,  G06N3/08 ,  G06T3/00 ,  G06N3/04 ,  G06T11/00 ,  G06K9/62

Abstract: A training method of a neural network for implementing image style transfer, an image processing method and an image processing device are disclosed. The training method includes: acquiring a first training input image and a second training input image; performing a style transfer process on the first training input image by the neural network to obtain a training output image; based on the first training input image, the second training input image and the training output image, calculating a loss value of parameters of the neural network through a loss function; and modifying the parameters of the neural network according to the loss value, where the loss function satisfies a predetermined condition, obtaining a trained neural network, where the loss function doesn't satisfy the predetermined condition, continuing to repeatedly perform above training process, the loss function including a weight-bias-ratio loss function.

公开(公告)号：US20210334642A1

公开(公告)日：2021-10-28

申请号：US16604410

申请日：2019-04-23

Applicant: BOE TECHNOLOGY GROUP CO., LTD.

Inventor： Pablo Navarrete Michelini ,  Dan Zhu ,  Hanwen Liu

IPC: G06N3/08 ,  G06N3/04 ,  G06K9/62

Abstract: The present disclosure generally relates to the field of deep learning technologies. An apparatus for generating a plurality of correlation images may include a feature extracting unit configured to receive a training image and extracting at least one or more of feature from the training image to generate a first feature image based on the training image; a normalizer configured to normalize the first feature image and generate a second feature image; and a shift correlating unit configured to perform a plurality of translational shifts on the second feature image to generate a plurality of shifted images, correlate each of the plurality of shifted images with the second feature image to generate the plurality of correlation images.

公开(公告)号：US20210097645A1

公开(公告)日：2021-04-01

申请号：US16465294

申请日：2018-12-17

Applicant: BOE TECHNOLOGY GROUP CO., LTD.

Inventor： Pablo Navarrete Michelini ,  Dan Zhu ,  Hanwen Liu

IPC: G06T3/40 ,  G06T5/50

Abstract: The present disclosure relates to an image processing method. The image processing method may include upscaling a feature image of an input image by an upscaling convolutional network to obtain a upscaled feature image; downscaling the upscaled feature image by a downscaling convolutional network to obtain a downscaled feature image; determining a residual image between the downscaled feature image and the feature image of the input image; upscaling the residual image between the downscaled feature image and the feature image of the input image to obtain an upscaled residual image; correcting the upscaled feature image using the upscaled residual image to obtain a corrected upscaled feature image; and generating a first super-resolution image based on the input image using the corrected upscaled feature image.

公开(公告)号：US20210012181A1

公开(公告)日：2021-01-14

申请号：US16626302

申请日：2019-07-22

Applicant: BOE Technology Group Co., Ltd.

Inventor： Dan Zhu ,  Lijie Zhang ,  Pablo Navarre Michelini ,  Hanwen Liu

IPC: G06N3/04 ,  G06N3/08 ,  G06T7/10 ,  G06F17/16

Abstract: A computer-implemented method of training a convolutional neural network configured to morph content features of an input image with style features of a style image is provided. The computer-implemented method includes selecting a training style image; extracting style features of the training style image; selecting a training content image; extracting content features of the training content image; processing the training content image through the convolutional neural network to generate a training output image including the content features of the training content image morphed with the style features of the training style image; extracting content features and style features of the training output image; computing a total loss; and tuning the convolutional neural network based on the total loss including a content loss, a style loss, and a regularization loss.

公开(公告)号：US10515435B2

公开(公告)日：2019-12-24

申请号：US15821095

申请日：2017-11-22

Applicant: BOE Technology Group Co., Ltd.

Inventor： Pablo Navarrete Michelini ,  Hanwen Liu ,  Xiaoyu Li

IPC: G06T3/40 ,  G06N3/08 ,  G06N3/04 ,  G06T1/20

Abstract: The disclosure discloses an apparatus for upscaling an image, a method for training the same, and a method for upscaling an image, where a convolutional neural network circuit obtains feature images of the image, a multiplexer upscales the image by integrating every n*n feature images of an input signal into a feature image with a resolution which is n times the resolution of a feature image of the image, where n is an integer greater than 1.

公开(公告)号：US20190139191A1

公开(公告)日：2019-05-09

申请号：US15921961

申请日：2018-03-15

Applicant: BOE Technology Group Co., Ltd.

Inventor： Hanwen Liu ,  Pablo Navarrete Michelini

IPC: G06T3/40 ,  G06T5/10 ,  G06T5/00 ,  G06K9/62

Abstract: The embodiments of the present disclosure provide an image processing method, and a processing device. The image processing method comprises: acquiring a first image including N components, where N is a positive integer greater than or equal to 1; and performing image conversion processing on the first image using a generative neural network, to output a first output image, wherein the generative neural network is trained using a Laplace transform function.

公开(公告)号：US12198302B2

公开(公告)日：2025-01-14

申请号：US17775340

申请日：2021-07-15

Applicant: BOE TECHNOLOGY GROUP CO., LTD.

Inventor： Ran Duan ,  Hanwen Liu ,  Yunhua Lu

IPC: G06T15/00 ,  G06T3/40 ,  G06T5/50 ,  G06T7/194 ,  G06T11/60 ,  G06V10/46

Abstract: An image processing method includes: acquiring a first image containing a target object; inputting the first image into an image processing model to obtain a second image, the second image being a mask image of the target object in the first image, a value for each pixel in the second image being in a range of 0 to 1, inclusive ([0, 1]), and the range of 0 to 1, inclusive ([0, 1]) indicating a degree of relation between each pixel in the second image and a pixel in the target object; fusing the first image and a background image according to the second image to obtain a fused image; and providing a first interface and displaying the fused image on the first interface.