知搜-全球专利检索

  1. Login
  2. Sign Up

Search Results

53 records (took 0.021 seconds)

    SYSTEMS AND METHODS FOR DEPTH ESTIMATION IN A VEHICLE
    31.
    发明申请
    SYSTEMS AND METHODS FOR DEPTH ESTIMATION IN A VEHICLE 有权

    公开(公告)号：US20220292289A1

    公开(公告)日：2022-09-15

    申请号：US17198954

    申请日：2021-03-11

    Applicant: GM GLOBAL TECHNOLOGY OPERATIONS LLC

    Inventor： Albert Shalumov ,  Michael Slutsky

    IPC: G06K9/00 ,  G06K9/62 ,  G06T7/13 ,  G06K9/72 ,  G06T7/55 ,  H04N7/18 ,  G06T5/00 ,  H04N5/247 ,  H04N5/232 ,  G06N3/04 ,  G06N3/08

    Abstract: Methods and system for training a neural network for depth estimation in a vehicle. The methods and systems receive respective training image data from at least two cameras. Fields of view of adjacent cameras of the at least two cameras partially overlap. The respective training image data is processed through a neural network providing depth data and semantic segmentation data as outputs. The neural network is trained based on a loss function. The loss function combines a plurality of loss terms including at least a semantic segmentation loss term and a panoramic loss term. The panoramic loss term includes a similarity measure regarding overlapping image patches of the respective image data that each correspond to a region of overlapping fields of view of the adjacent cameras. The semantic segmentation loss term quantifies a difference between ground truth semantic segmentation data and the semantic segmentation data output from the neural network.

    USING EPIPOLAR REPROJECTION FOR VIRTUAL VIEW PERSPECTIVE CHANGE
    32.
    发明申请
    USING EPIPOLAR REPROJECTION FOR VIRTUAL VIEW PERSPECTIVE CHANGE 有权

    公开(公告)号：US20220284660A1

    公开(公告)日：2022-09-08

    申请号：US17189877

    申请日：2021-03-02

    Applicant: GM GLOBAL TECHNOLOGY OPERATIONS LLC

    Inventor： Michael Slutsky ,  Albert Shalumov

    IPC: G06T15/20 ,  G06T7/70 ,  G06T7/50 ,  H04N7/18

    Abstract: Systems and methods for generating a virtual view of a virtual camera based on an input scene are described. A capturing device typically includes a physical camera and a depth sensor and captures an input scene. A controller determines an actual pose of the capturing device and a desired pose of the virtual camera for showing the virtual view. The controller defines an epipolar geometry between the actual pose of the capturing device and the desired pose of the virtual camera. The controller generates an output image for the virtual camera based on an epipolar relation between the actual pose of the capturing device, the input scene, and the desired pose of the virtual camera.

    Systems and methods for training a non-blind image deblurring module
    33.
    发明授权
    Systems and methods for training a non-blind image deblurring module 有权

    公开(公告)号：US11354784B2

    公开(公告)日：2022-06-07

    申请号：US16806135

    申请日：2020-03-02

    Applicant: GM GLOBAL TECHNOLOGY OPERATIONS LLC

    Inventor： Michael Slutsky

    IPC: G06T5/00 ,  G06N3/08 ,  B60R1/00 ,  G06K9/62

    Abstract: Methods and systems for training a non-blind deblurring module are disclosed. Unblurred test images and blurred test images are received, wherein each of the blurred test images is related to a corresponding one of the unblurred test images by a blur kernel term and a noise term. A regularized deconvolution sub-module and a convolutional neural network are jointly trained by adjusting a regularization parameter of a regularized deconvolution function and weights of a convolution neural network in order to minimize a cost function representative of a difference between each deblurred output image and a corresponding one of the unblurred test images.

    SYSTEMS AND METHODS FOR TRAINING A NON-BLIND IMAGE DEBLURRING MODULE
    34.
    发明申请
    SYSTEMS AND METHODS FOR TRAINING A NON-BLIND IMAGE DEBLURRING MODULE 有权

    公开(公告)号：US20210272248A1

    公开(公告)日：2021-09-02

    申请号：US16806135

    申请日：2020-03-02

    Applicant: GM GLOBAL TECHNOLOGY OPERATIONS LLC

    Inventor： Michael Slutsky

    IPC: G06T5/00 ,  G06N3/08 ,  G06K9/62 ,  B60R1/00

    Abstract: Methods and systems for training a non-blind deblurring module are disclosed. Unblurred test images and blurred test images are received, wherein each of the blurred test images is related to a corresponding one of the unblurred test images by a blur kernel term and a noise term. A regularized deconvolution sub-module and a convolutional neural network are jointly trained by adjusting a regularization parameter of a regularized deconvolution function and weights of a convolution neural network in order to minimize a cost function representative of a difference between each deblurred output image and a corresponding one of the unblurred test images.

    NATURAL SURROUND VIEW
    35.
    发明申请
    NATURAL SURROUND VIEW 有权

    公开(公告)号：US20210264166A1

    公开(公告)日：2021-08-26

    申请号：US16801587

    申请日：2020-02-26

    Applicant: GM GLOBAL TECHNOLOGY OPERATIONS LLC

    Inventor： Michael Slutsky ,  Daniel Kigli

    IPC: G06K9/00 ,  G06K9/62 ,  G06T7/536 ,  G06N3/08

    Abstract: In various embodiments, methods and systems are provided for processing camera data from a camera system associated with a vehicle. In one embodiment, a method includes: storing a plurality of photorealistic scenes of an environment; training, by a processor, a machine learning model to produce a surround view approximating a ground truth surround view using the plurality of photorealistic scenes as training data; and processing, by a processor, the camera data from the camera system associated with the vehicle based on the trained machine learning model to produce a surround view of an environment of the vehicle.

    GENERATION OF COMPOSITE IMAGES USING INTERMEDIATE IMAGE SURFACES
    36.
    发明申请
    GENERATION OF COMPOSITE IMAGES USING INTERMEDIATE IMAGE SURFACES 有权

    公开(公告)号：US20210158493A1

    公开(公告)日：2021-05-27

    申请号：US16691198

    申请日：2019-11-21

    Applicant: GM GLOBAL TECHNOLOGY OPERATIONS LLC

    Inventor： Michael Slutsky ,  Yael Gefen ,  Lior Stein

    IPC: G06T5/00 ,  G06T5/50

    Abstract: A system for processing images includes a receiving module configured to receive a plurality of images generated by one or more imaging devices, the plurality of images including a first image taken from a first location and orientation, and a second image taken from a second location and orientation. The system also includes an image analysis module configured to generate a composite image on a target image surface based on at least the first image and the second image. The image analysis module is configured to perform steps that include selecting a planar intermediate image surface, projecting the first image and the second image onto the intermediate image surface, combining the projected first image and the projected second image at the intermediate image surface to generate an intermediate image, and projecting the intermediate image onto the target image surface to generate the composite image.

    VEHICLE IMAGING SYSTEM AND METHOD FOR A PARKING SOLUTION
    37.
    发明申请
    VEHICLE IMAGING SYSTEM AND METHOD FOR A PARKING SOLUTION 审中-公开

    公开(公告)号：US20200282909A1

    公开(公告)日：2020-09-10

    申请号：US16295911

    申请日：2019-03-07

    Applicant: GM GLOBAL TECHNOLOGY OPERATIONS LLC

    Inventor： Nicky Zimmerman ,  Yael Shmueli Friedland ,  Michael Slutsky

    IPC: B60R1/00

    Abstract: A vehicle imaging system and method for providing a user with an easy to use vehicle parking solution that displays an integrated and intuitive backup camera view, such as a first-person composite camera view. The first-person composite camera view may include composite image data from a plurality of cameras mounted around the vehicle that has been joined or stitched together, as well as augmented graphics with computer-generated simulations of parts of the vehicle that provide the user with intuitive information concerning the point-of-view being displayed. The point-of-view of the first-person composite camera view is that of an observer located within the vehicle, and is designed to emulate the point-of-view of a driver. It is also possible to provide a direction indicator that allows the user to engage a touch screen display and manually change the direction of the first-person composite camera view so that the user can intuitively explore the vehicle surroundings.

    Generation of artificial color images from narrow spectral band data aboard a camera-equipped vehicle
    38.
    发明授权
    Generation of artificial color images from narrow spectral band data aboard a camera-equipped vehicle 有权

    公开(公告)号：US12223741B2

    公开(公告)日：2025-02-11

    申请号：US17557448

    申请日：2021-12-21

    Applicant: GM GLOBAL TECHNOLOGY OPERATIONS LLC

    Inventor： Tzvi Philipp ,  Eran Kishon ,  Michael Slutsky

    IPC: G06V20/58 ,  B60W50/14 ,  G02B27/01 ,  G06V10/25

    Abstract: A visual perception system includes a scanning camera, color sensor with color filter array (CFA), and a classifier node. The camera captures full color pixel images of a target object, e.g., a traffic light, and processes the pixel images through a narrow band pass filter (BPF), such that the narrow BPF outputs monochromatic images of the target object. The color sensor and CFA receive the monochromatic images. The color sensor has at least three color channels each corresponding to different colors of spectral data in the monochromatic images. The classifier node uses a predetermined classification decision tree to classify constituent pixels of the monochromatic images into different color bins as a corresponding color of interest. The color of interest may be used to perform a control action, e.g., via an automated driver assist system (ADAS) control unit or an indicator device.

Patent Agency Ranking