Patents by Inventor Weihong Deng

Weihong Deng has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).

  • Patent number: 12165398
    Abstract: The present disclosure relates to training method and apparatus for an object recognition model. There provides a training sample optimization apparatus for a neural network model for object recognition, comprising: for each training sample in a training sample database, a fluctuation determination unit configured to determine a fluctuation of model prediction of the training sample relative to a corresponding labeled identity of the training sample in a case of training the neural network model; and an optimization unit configured to determine whether the training sample can be used for training of the neural network model in the next training epoch, based on the fluctuation of the training sample.
    Type: Grant
    Filed: December 6, 2021
    Date of Patent: December 10, 2024
    Assignee: Canon Kabushiki Kaisha
    Inventors: Dongyue Zhao, Dongchao Wen, Xian Li, Weihong Deng, Jiani Hu
  • Patent number: 12026974
    Abstract: The present invention relates to method and apparatus for training a neural network for object recognition. A training method which includes inputting a training image set containing an object to be recognized, dividing the image samples in the training image set into simple samples and hard samples, for each kind of the image sample and the variation image sample, performing, a transitive transfer, calculating a distillation loss of the transferred student feature of the image sample relative to a teacher feature extracted from corresponding image sample of the other kind, classifying, the image sample, and calculating a classification loss of the image sample, calculating a total loss related to the training image set, and updating parameters of the neural network according to the calculated total loss.
    Type: Grant
    Filed: November 4, 2021
    Date of Patent: July 2, 2024
    Assignee: Canon Kabushiki Kaisha
    Inventors: Dongyue Zhao, Dongchao Wen, Xian Li, Weihong Deng, Jiani Hu
  • Publication number: 20220180627
    Abstract: The present disclosure relates to training method and apparatus for an object recognition model. There provides a training sample optimization apparatus for a neural network model for object recognition, comprising: for each training sample in a training sample database, a fluctuation determination unit configured to determine a fluctuation of model prediction of the training sample relative to a corresponding labeled identity of the training sample in a case of training the neural network model; and an optimization unit configured to determine whether the training sample can be used for training of the neural network model in the next training epoch, based on the fluctuation of the training sample.
    Type: Application
    Filed: December 6, 2021
    Publication date: June 9, 2022
    Inventors: Dongyue Zhao, Dongchao Wen, Xian Li, Weihong Deng, Jiani Hu
  • Publication number: 20220138454
    Abstract: The present invention relates to method and apparatus for training a neural network for object recognition. A training method which includes inputting a training image set containing an object to be recognized, dividing the image samples in the training image set into simple samples and hard samples, for each kind of the image sample and the variation image sample, performing, a transitive transfer, calculating a distillation loss of the transferred student feature of the image sample relative to a teacher feature extracted from corresponding image sample of the other kind, classifying, the image sample, and calculating a classification loss of the image sample, calculating a total loss related to the training image set, and updating parameters of the neural network according to the calculated total loss.
    Type: Application
    Filed: November 4, 2021
    Publication date: May 5, 2022
    Inventors: Dongyue Zhao, Dongchao Wen, Xian Li, Weihong Deng, Jiani Hu
  • Patent number: 11200451
    Abstract: A template determining apparatus including an attribute distribution determination unit configured to determine a distribution of a specific attribute in a plurality of images; and a template determination unit configured to adaptatively determine a template set from the plurality of images according to the determined distribution of the specific attribute of the plurality of images. Where the determined template set will be used for image normalization.
    Type: Grant
    Filed: November 11, 2019
    Date of Patent: December 14, 2021
    Assignee: Canon Kabushiki Kaisha
    Inventors: Yaohai Huang, Jianteng Peng, Weihong Deng, Jiani Hu
  • Publication number: 20210241097
    Abstract: A training method and device for an object recognition model. An apparatus for optimizing a neural network model for object recognition, including a loss determination unit configured to determine loss data for features extracted from a training image set using the neural network model and a loss function with a weight function, and an updating unit configured to perform an updating operation on parameters of the neural network model based on the loss data and an updating function, wherein the updating function is derived based on the loss function with the weight function of the neural network model, and the weight function and the loss function change monotonically in a specific value interval in the same direction.
    Type: Application
    Filed: November 4, 2020
    Publication date: August 5, 2021
    Inventors: Dongyue Zhao, Dongchao Wen, Xian Li, Weihong Deng, Jiani Hu
  • Publication number: 20200151502
    Abstract: A template determining apparatus including an attribute distribution determination unit configured to determine a distribution of a specific attribute in a plurality of images; and a template determination unit configured to adaptatively determine a template set from the plurality of images according to the determined distribution of the specific attribute of the plurality of images. Where the determined template set will be used for image normalization.
    Type: Application
    Filed: November 11, 2019
    Publication date: May 14, 2020
    Inventors: Yaohai Huang, Jianteng Peng, Weihong Deng, Jiani Hu