利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

58 条记录 (耗时 0.028 秒)

    NETWORK ARCHITECTURE FOR MONOCULAR DEPTH ESTIMATION AND OBJECT DETECTION
    6.
    发明申请
    NETWORK ARCHITECTURE FOR MONOCULAR DEPTH ESTIMATION AND OBJECT DETECTION 有权

    公开(公告)号：US20220301202A1

    公开(公告)日：2022-09-22

    申请号：US17333537

    申请日：2021-05-28

    申请人： Toyota Research Institute, Inc.

    发明人： Dennis Park ,  Rares A. Ambrus ,  Vitor Guizilini ,  Jie Li ,  Adrien David Gaidon

    IPC分类号： G06T7/50 ,  G06N3/08

    摘要： System, methods, and other embodiments described herein relate to performing depth estimation and object detection using a common network architecture. In one embodiment, a method includes generating, using a backbone of a combined network, a feature map at multiple scales from an input image. The method includes decoding, using a top-down pathway of the combined network, the feature map to provide features at the multiple scales. The method includes generating, using a head of the combined network, a depth map from the features for a scene depicted in the input image, and bounding boxes identifying objects in the input image.

    SYSTEMS AND METHODS FOR TRAJECTORY FORECASTING ACCORDING TO SEMANTIC CATEGORY UNCERTAINTY
    7.
    发明申请
    SYSTEMS AND METHODS FOR TRAJECTORY FORECASTING ACCORDING TO SEMANTIC CATEGORY UNCERTAINTY 有权

    公开(公告)号：US20220180170A1

    公开(公告)日：2022-06-09

    申请号：US17112292

    申请日：2020-12-04

    申请人： Toyota Research Institute, Inc.

    发明人： Boris Ivanovic ,  Kuan-Hui Lee ,  Jie Li ,  Adrien David Gaidon ,  Pavel Tokmakov

    IPC分类号： G06N3/08 ,  G06N3/04 ,  B60W30/095

    摘要： System, methods, and other embodiments described herein relate to improving trajectory forecasting in a device. In one embodiment, a method includes, in response to receiving sensor data about a surrounding environment of the device, identifying an object from the sensor data that is present in the surrounding environment. The method includes determining category probabilities for the object, the category probabilities indicating semantic classes for classifying the object and probabilities that the object belongs to the semantic classes. The method includes forecasting trajectories for the object based, at least in part, on the category probabilities and the sensor data. The method includes controlling the device according to the trajectories.

    SYSTEM AND METHOD FOR TRAINING OF A DETECTOR MODEL TO OUTPUT AN INSTANCE IDENTIFIER INDICATING OBJECT CONSISTENCY ALONG THE TEMPORAL AXIS
    8.
    发明申请
    SYSTEM AND METHOD FOR TRAINING OF A DETECTOR MODEL TO OUTPUT AN INSTANCE IDENTIFIER INDICATING OBJECT CONSISTENCY ALONG THE TEMPORAL AXIS 有权

    公开(公告)号：US20220036126A1

    公开(公告)日：2022-02-03

    申请号：US16943393

    申请日：2020-07-30

    申请人： Toyota Research Institute, Inc.

    发明人： Adrien David Gaidon ,  Jie Li

    IPC分类号： G06K9/62 ,  G06N20/00 ,  G06T7/70

    摘要： A detector system having a detector model includes one or more processor(s) and a memory. The memory includes an image acquisition module, a training module, and a label propagating module. The modules cause the processor(s) to obtain a first training set, train the detector model using the first training set and a first loss function, label propagate a second training set by the detector model after the detector model is trained with the first training set, and train the detector model using the first training set, the second training set, the first loss function, and a discriminative loss function. The detector model is trained through an intermediate multidimensional feature predicted at each pixel location of the one or more objects of the first training set and the second training set. The intermediate multidimensional feature being an instance identifier expressing the temporal consistency of objects along the temporal axis.