知搜-全球专利检索

  1. Login
  2. Sign Up

Search Results

68 records (took 0.022 seconds)

    DIRECT THIN BOUNDARY PREDICTION
    21.
    发明申请
    DIRECT THIN BOUNDARY PREDICTION 审中-公开

    公开(公告)号：US20200065973A1

    公开(公告)日：2020-02-27

    申请号：US16545158

    申请日：2019-08-20

    Applicant: Apple Inc.

    Inventor： Amit Kumar KC ,  Daniel Ulbricht

    IPC: G06T7/13 ,  G06N3/08

    Abstract: In some implementations a neural network is trained to perform to directly predict thin boundaries of objects in images based on image characteristics. A neural network can be trained to predict thin boundaries of objects without requiring subsequent computations to reduce the thickness of the boundary prediction. Instead, the network is trained to make the predicted boundaries thin by effectively suppressing non-maximum values in normal directions along what might otherwise be a thick predicted boundary. To do so, the neural network can be trained to determine normal directions and suppress non-maximum values based on those determined normal directions.

    Pixel Classification Techniques
    22.
    发明申请
    Pixel Classification Techniques 审中-公开

    公开(公告)号：US20180018541A1

    公开(公告)日：2018-01-18

    申请号：US15213018

    申请日：2016-07-18

    Applicant: Apple Inc.

    Inventor： Lejing Wang ,  Daniel Ulbricht ,  Lorenz Kern ,  Sebastian Lieberknecht ,  Tobias Holl

    IPC: G06K9/62 ,  H04N5/225

    CPC classification number: G06K9/6267 ,  G06K9/6298 ,  G06T7/11 ,  G06T7/136 ,  G06T7/143 ,  G06T7/194 ,  G06T2207/20012 ,  H04N5/2253 ,  H04N5/2254

    Abstract: Systems, methods, and computer readable media to categorize a pixel (or other element) in an image into one of a number of different categories are described. In general, techniques are disclosed for using properties (e.g., statistics) of the regions being categorized to determine the appropriate size of window around a target pixel (element) and, when necessary, the manner in which the window may be changed if the current size is inappropriate. More particularly, adaptive window size selection techniques are disclosed for use when categorizing an image's pixels into one of two categories (e.g., black or white). Statistics of the selected region may be cascaded to determine whether the current evaluation window is acceptable and, if it is not, an appropriate factor by which to change the currently selected window's size

    OBJECT RELATIONSHIP ESTIMATION FROM A 3D SEMANTIC MESH
    23.
    发明申请
    OBJECT RELATIONSHIP ESTIMATION FROM A 3D SEMANTIC MESH 有权

    公开(公告)号：US20250068781A1

    公开(公告)日：2025-02-27

    申请号：US18945802

    申请日：2024-11-13

    Applicant: APPLE INC.

    Inventor： Angela Blechschmidt ,  Daniel Ulbricht ,  Omar Elafifi

    IPC: G06F30/10 ,  G06N3/08 ,  G06N20/00

    Abstract: Implementations disclosed herein provide systems and methods that determine relationships between objects based on an original semantic mesh of vertices and faces that represent the 3D geometry of a physical environment. Such an original semantic mesh may be generated and used to provide input to a machine learning model that estimates relationships between the objects in the physical environment. For example, the machine learning model may output a graph of nodes and edges indicating that a vase is on top of a table or that a particular instance of a vase, V1, is on top of a particular instance of a table, T1.

    Electronic asset management
    24.
    发明授权
    Electronic asset management 有权

    公开(公告)号：US12094019B1

    公开(公告)日：2024-09-17

    申请号：US16944812

    申请日：2020-07-31

    Applicant: Apple Inc.

    Inventor： Jack R. Greasley ,  Daniel Ulbricht ,  Eshan Verma

    IPC: G06Q50/18 ,  G06F16/23

    CPC classification number: G06Q50/184 ,  G06F16/23

    Abstract: Various implementations manage an electronic asset by creating a representation of an electronic asset and its variants. This may be accomplished by identifying variants of an electronic asset, identifying a portion of a feature space associated with the asset and variants, and providing a representation corresponding to that portion of feature space. A fixed function classifier may be used to determine the points in the feature space for the electronic asset and its variants. The set of points produced for an asset and its variants using such a fixed function classifier will be near one another in feature space. Moreover, the area around such points will also represent points for other similar variations of the asset and thus, the portion of the feature space around the points can be considered the area of ownership for the electronic asset, e.g., it defines a boundary of what the creator is asserting is his or her creation.

    CONTENT EVENT MAPPING
    27.
    发明申请
    CONTENT EVENT MAPPING 有权

    公开(公告)号：US20220060802A1

    公开(公告)日：2022-02-24

    申请号：US17275038

    申请日：2019-09-24

    Applicant: Apple Inc.

    Inventor： Ian M. Richter ,  Daniel Ulbricht ,  Eshan Verma

    IPC: H04N21/81 ,  G06T19/00 ,  G06K9/00 ,  H04N21/44 ,  H04N21/845

    Abstract: In one implementation, consumption of media content (such as video, audio, or text) is supplemented with an immersive synthesized reality (SR) map based on the media content. In various implementations described herein, the SR map includes a plurality of SR environment representations which, when selected by a user, cause display of a corresponding SR environment.

    Environment-based application presentation
    28.
    发明授权
    Environment-based application presentation 有权

    公开(公告)号：US11074451B2

    公开(公告)日：2021-07-27

    申请号：US16833364

    申请日：2020-03-27

    Applicant: Apple Inc.

    Inventor： Peter Meier ,  Daniel Ulbricht

    IPC: G06K9/00 ,  G06T19/00

    Abstract: In an exemplary process for providing content in an augmented reality environment, image data correspond to a physical environment are obtained. Based on the image data, predefined entities of the plurality of predefined entities in the physical environment are identified using classifiers corresponding to predefined entities. Based on the one or more of the identified predefined entities, a geometric layout of the physical environment is determined. Based on the geometric layout, an area corresponding to a particular entity is determined. The particular entity corresponds to one or more identified predefined entities. Based on the area corresponding to the particular entity, the particular entity in the physical environment is identified using classifiers corresponding to the determined area. Based on the identified particular entity, a type of the physical environment is determined. Based on the type of the physical environment, virtual-reality objects are displayed corresponding to a representation of the physical environment.

    Method and device for pixel-level object segmentation
    29.
    发明授权
    Method and device for pixel-level object segmentation 有权

    公开(公告)号：US11048977B1

    公开(公告)日：2021-06-29

    申请号：US16580294

    申请日：2019-09-24

    Applicant: Apple Inc.

    Inventor： Daniel Ulbricht ,  Mohammed Haris Baig ,  Amit Kumar K C

    IPC: G06K9/62 ,  G06T7/10

    Abstract: In one implementation, a method of training a type-agnostic object segmentation system is performed in a type-agnostic object segmentation system including one or more processors, and a non-transitory memory. The method includes obtaining a data set; generating a respective embedding vector for each of a plurality of pixels of the image frame; mapping a plurality of embedding vector representations into a dataspace defined by a dimensionality of the plurality of embedding vector representations; comparing the at least one object instance representation candidate against the respective segmentation mask; and adjusting the type-agnostic object segmentation system in order to satisfy an error threshold across the plurality of image data frames according to a determination that the at least one object instance representation candidate and the respective segmentation mask differ by a threshold value.

Patent Agency Ranking