Patents by Inventor Bingzhe Wu

Bingzhe Wu 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).

  • Generator-to-classifier framework for object classification
    Patent number: 10740895
    Abstract: A computer implemented method, computer system and computer program product are provided for data classification. The method includes receiving original data, wherein the original data includes at least one object in a first condition. The method also includes receiving generated data from a generator based on the original data, wherein the generated data includes the at least one object in a second condition, the generator trained by training data of the first condition and training data of the second condition. The method further includes determining a classification of the at least one object with a classifier based on the original data and the generated data, the classifier trained by labeled data of the first condition and more training data of the second condition that is generated based on the labeled data by the trained generator, wherein the labeled data includes the at least one object.
    Type: Grant
    Filed: June 25, 2018
    Date of Patent: August 11, 2020
    Assignee: International Business Machines Corporation
    Inventors: Shiwan Zhao, Xiaolu Zhang, Shi Lei Zhang, Bingzhe Wu
  • GENERATOR-TO-CLASSIFIER FRAMEWORK FOR OBJECT CLASSIFICATION
    Publication number: 20190392576
    Abstract: A computer implemented method, computer system and computer program product are provided for data classification. The method includes receiving original data, wherein the original data includes at least one object in a first condition. The method also includes receiving generated data from a generator based on the original data, wherein the generated data includes the at least one object in a second condition, the generator trained by training data of the first condition and training data of the second condition. The method further includes determining a classification of the at least one object with a classifier based on the original data and the generated data, the classifier trained by labeled data of the first condition and more training data of the second condition that is generated based on the labeled data by the trained generator, wherein the labeled data includes the at least one object.
    Type: Application
    Filed: June 25, 2018
    Publication date: December 26, 2019
    Inventors: Shiwan Zhao, Xiaolu Zhang, Shi Lei Zhang, Bingzhe Wu