Patents by Inventor Dongzhe Yang

Dongzhe Yang 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: 11836223
    Abstract: The disclosed computer-implemented method may include collecting a set of labels that label polygons within a training set of images as architectural structures. The method may also include creating a set of noisy labels with a predetermined degree of noise by distorting boundaries of a number of the polygons within the training set of images. Additionally, the method may include simultaneously training two neural networks by applying a co-teaching method to learn from the set of noisy labels. The method may also include extracting a preferential list of training data based on the two trained neural networks. Furthermore, the method may include training a machine learning model with the preferential list of training data. Finally, the method may include identifying one or more building footprints in a target image using the trained machine learning model. Various other methods, systems, and computer-readable media are also disclosed.
    Type: Grant
    Filed: June 17, 2021
    Date of Patent: December 5, 2023
    Assignee: Meta Platforms, Inc.
    Inventors: Li Chen, Purvi Goel, Ilknur Kaynar Kabul, David Dongzhe Yang
  • Publication number: 20220156526
    Abstract: The disclosed computer-implemented method may include collecting a set of labels that label polygons within a training set of images as architectural structures. The method may also include creating a set of noisy labels with a predetermined degree of noise by distorting boundaries of a number of the polygons within the training set of images. Additionally, the method may include simultaneously training two neural networks by applying a co-teaching method to learn from the set of noisy labels. The method may also include extracting a preferential list of training data based on the two trained neural networks. Furthermore, the method may include training a machine learning model with the preferential list of training data. Finally, the method may include identifying one or more building footprints in a target image using the trained machine learning model. Various other methods, systems, and computer-readable media are also disclosed.
    Type: Application
    Filed: June 17, 2021
    Publication date: May 19, 2022
    Inventors: Li Chen, Purvi Goel, Ilknur Kaynar Kabul, David Dongzhe Yang
  • Patent number: 8046202
    Abstract: The present invention provides for converting a user-specified element of a graphical model into at least one intermediate representation that represents the user-specified element. An intermediate representation of the present invention may support simulation, propagation, and/or code generation. An element for a graphical model is provided, and the element and/or the graphical model are converted into an intermediate representation that may represent algorithmic aspects of the element or that is capable of being simulated and/or capable allowing propagation. The present invention allows a user to provide a single code for a user-specified element that may be used to support simulation, propagation and/or code generation of an element or graphical model.
    Type: Grant
    Filed: July 17, 2006
    Date of Patent: October 25, 2011
    Assignee: The MathWorks, Inc.
    Inventors: Dongzhe Yang, Robert O. Aberg, Pieter J. Mosterman
  • Patent number: 7805398
    Abstract: The present invention provides a method and system for registering rules for attributes of an entity with an entity validation and inference unit. The entity validation and inference unit can further validate input values for an entity and/or infer attributes of an entity using the associated registered rules. The entity validation and inference unit can also provide information regarding attributes, relations, and behaviors of an entity using the associated registered rules in response to a query. The present invention also provides a mechanism for a user to query an entity or the entity validation and inference unit regarding valid values for an entity. The mechanism also allows the user to query the registered rule associated with the entity.
    Type: Grant
    Filed: October 1, 2009
    Date of Patent: September 28, 2010
    Assignee: The MathWorks, Inc.
    Inventor: Dongzhe Yang
  • Patent number: 7610256
    Abstract: The present invention provides a method and system for registering rules for attributes of an entity with an entity validation and inference unit. The entity validation and inference unit can further validate input values for an entity and/or infer attributes of an entity using the associated registered rules. The entity validation and inference unit can also provide information regarding attributes, relations, and behaviors of an entity using the associated registered rules in response to a query. The present invention also provides a mechanism for a user to query an entity or the entity validation and inference unit regarding valid values for an entity. The mechanism also allows the user to query the registered rule associated with the entity.
    Type: Grant
    Filed: August 20, 2007
    Date of Patent: October 27, 2009
    Assignee: The MathWorks, Inc.
    Inventor: Dongzhe Yang
  • Patent number: 7472103
    Abstract: The present invention provides a method and system for registering rules for attributes of an entity with an entity validation and inference unit. The entity validation and inference unit can further validate input values for an entity and/or infer attributes of an entity using the associated registered rules. The entity validation and inference unit can also provide information regarding attributes, relations, and behaviors of an entity using the associated registered rules in response to a query. The present invention also provides a mechanism for a user to query an entity or the entity validation and inference unit regarding valid values for an entity. The mechanism also allows the user to query the registered rule associated with the entity.
    Type: Grant
    Filed: December 23, 2005
    Date of Patent: December 30, 2008
    Assignee: The MathWorks, Inc.
    Inventor: Dongzhe Yang