利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

8 条记录 (耗时 0.025 秒)

    Training methods for deep networks
    2.
    发明授权
    Training methods for deep networks 有权

    公开(公告)号：US11113526B2

    公开(公告)日：2021-09-07

    申请号：US16570813

    申请日：2019-09-13

    申请人： TOYOTA RESEARCH INSTITUTE, INC.

    发明人： Kevin Stone ,  Krishna Shankar ,  Michael Laskey

    IPC分类号： G06T7/73 ,  G06K9/62 ,  B25J9/16 ,  G06K9/00 ,  G06T7/55 ,  G06N3/08 ,  G06T19/20

    摘要： A method for training a deep neural network of a robotic device is described. The method includes constructing a 3D model using images captured via a 3D camera of the robotic device in a training environment. The method also includes generating pairs of 3D images from the 3D model by artificially adjusting parameters of the training environment to form manipulated images using the deep neural network. The method further includes processing the pairs of 3D images to form a reference image including embedded descriptors of common objects between the pairs of 3D images. The method also includes using the reference image from training of the neural network to determine correlations to identify detected objects in future images.

    EMBEDDINGS + SVM FOR TEACHING TRAVERSABILITY
    4.
    发明申请
    EMBEDDINGS + SVM FOR TEACHING TRAVERSABILITY 有权

    公开(公告)号：US20210081724A1

    公开(公告)日：2021-03-18

    申请号：US16697290

    申请日：2019-11-27

    申请人： Toyota Research Institute, Inc.

    发明人： Jeremy Ma ,  Krishna Shankar ,  Kevin Stone

    IPC分类号： G06K9/62 ,  G06N3/04 ,  G06N3/08 ,  G06N20/10 ,  G06N5/04

    摘要： A system includes a memory module configured to store image data captured by a camera and an electronic controller communicatively coupled to the memory module. The electronic controller is configured to receive image data captured by the camera, implement a neural network trained to predict a drivable portion in the image data of an environment. The neural network predicts the drivable portion in the image data of the environment. The electronic controller is configured to implement a support vector machine. The support vector machine determines whether the predicted drivable portion of the environment output by the neural network is classified as drivable based on a hyperplane of the support vector machine and output an indication of the drivable portion of the environment.

    Embeddings + SVM for teaching traversability
    5.
    发明授权
    Embeddings + SVM for teaching traversability 有权

    公开(公告)号：US11586861B2

    公开(公告)日：2023-02-21

    申请号：US16697290

    申请日：2019-11-27

    申请人： Toyota Research Institute, Inc.

    发明人： Jeremy Ma ,  Krishna Shankar ,  Kevin Stone

    IPC分类号： G06K9/62 ,  G06N5/04 ,  G06N3/08 ,  G06N20/10 ,  G06N3/04

    摘要： A system includes a memory module configured to store image data captured by a camera and an electronic controller communicatively coupled to the memory module. The electronic controller is configured to receive image data captured by the camera, implement a neural network trained to predict a drivable portion in the image data of an environment. The neural network predicts the drivable portion in the image data of the environment. The electronic controller is configured to implement a support vector machine. The support vector machine determines whether the predicted drivable portion of the environment output by the neural network is classified as drivable based on a hyperplane of the support vector machine and output an indication of the drivable portion of the environment.

    Unknown object identification for robotic device
    7.
    发明授权
    Unknown object identification for robotic device 有权

    公开(公告)号：US11328170B2

    公开(公告)日：2022-05-10

    申请号：US16795242

    申请日：2020-02-19

    申请人： TOYOTA RESEARCH INSTITUTE, INC.

    发明人： Krishna Shankar

    IPC分类号： G06K9/62 ,  G06V10/56 ,  G06V20/10

    摘要： A method for identifying objects includes generating a feature vector of an unknown object from an image of an environment. The method also includes comparing the feature vector of the unknown object to feature vectors of known objects. The method further includes determining whether a similarity between the feature vector of the unknown object and the feature vector of one of the known objects satisfies a threshold. Furthermore, the method includes identifying the unknown object based on the determination.