利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

20 条记录 (耗时 0.02 秒)

    Predictive parking due to weather
    2.
    发明授权
    Predictive parking due to weather 有权

    公开(公告)号：US11479239B2

    公开(公告)日：2022-10-25

    申请号：US17097727

    申请日：2020-11-13

    申请人： TOYOTA RESEARCH INSTITUTE, INC.

    发明人： Nikos Arechiga-Gonzalez ,  Soonho Kong ,  Jonathan Decastro ,  Frank Permenter ,  Dennis Park

    IPC分类号： H04N5/44 ,  B60W30/06 ,  B60W40/02 ,  B60W50/14 ,  G08G1/14

    摘要： Aspects of the disclosure provide a parking method and system for a vehicle. For example, the parking system can include a processor and an output circuit. As another example, the parking system can include the processor and a compelling unit. The processor can be configured to determine whether a first weather condition, such as snowing, raining, icing and hailing, at a first parking location meets a first criterion, and identify, when the first weather condition meets the first criterion, a second parking location at which a second weather condition meets a second criterion different from the first criterion. The output unit can be configured to output an alert to the vehicle parked at the first parking location indicating that the vehicle is to be moved to the second parking location. The compelling unit can be configured to compel the vehicle to move to the second parking location.

    Systems and methods for panoptic image segmentation
    6.
    发明授权
    Systems and methods for panoptic image segmentation 有权

    公开(公告)号：US11501525B2

    公开(公告)日：2022-11-15

    申请号：US16843026

    申请日：2020-04-08

    申请人： Toyota Research Institute, Inc.

    发明人： Rui Hou ,  Jie Li ,  Vitor Guizilini ,  Adrien David Gaidon ,  Dennis Park ,  Arjun Bhargava

    IPC分类号： G06V20/10 ,  G06T7/10 ,  G06K9/62 ,  G06V10/94

    摘要： Systems and methods for panoptic image segmentation are disclosed herein. One embodiment performs semantic segmentation and object detection on an input image, wherein the object detection generates a plurality of bounding boxes associated with an object in the input image; selects a query bounding box from among the plurality of bounding boxes; maps at least one of the bounding boxes in the plurality of bounding boxes other than the query bounding box to the query bounding box based on similarity between the at least one of the bounding boxes and the query bounding box to generate a mask assignment for the object, the mask assignment defining a contour of the object; compares the mask assignment with results of the semantic segmentation to produce a refined mask assignment for the object; and outputs a panoptic segmentation of the input image that includes the refined mask assignment for the object.

    SYSTEM AND METHOD FOR TRAINING A MULTI-TASK MODEL
    7.
    发明申请
    SYSTEM AND METHOD FOR TRAINING A MULTI-TASK MODEL 有权

    公开(公告)号：US20220300851A1

    公开(公告)日：2022-09-22

    申请号：US17205709

    申请日：2021-03-18

    申请人： Toyota Research Institute, Inc.

    发明人： Dennis Park ,  Adrien David Gaidon

    IPC分类号： G06N20/00 ,  G06N5/04

    摘要： A system for training a multi-task model includes a processor and a memory in communication with the processor. The memory has a multi-task training module having instructions that, when executed by the processor, causes the processor to provide simulation training data having a plurality of samples to a multi-task model capable of performing at least a first task and a second task using at least one shared. The training module further causes the processor to determine a first value (gradience or loss) for the first task and a second value (gradience or loss) for a second task using the simulation training data and the at least one shared parameter, determine a task induced variance between the first value and the second value, and iteratively adjust the at least one shared parameter to reduce the task induced variance.

    NETWORK ARCHITECTURE FOR MONOCULAR DEPTH ESTIMATION AND OBJECT DETECTION
    8.
    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.

    System and method for training a multi-task model
    9.
    发明授权
    System and method for training a multi-task model 有权

    公开(公告)号：US12086695B2

    公开(公告)日：2024-09-10

    申请号：US17205709

    申请日：2021-03-18

    申请人： Toyota Research Institute, Inc.

    发明人： Dennis Park ,  Adrien David Gaidon

    IPC分类号： G06N20/00 ,  G06N5/04

    CPC分类号： G06N20/00 ,  G06N5/04

    摘要： A system for training a multi-task model includes a processor and a memory in communication with the processor. The memory has a multi-task training module having instructions that, when executed by the processor, causes the processor to provide simulation training data having a plurality of samples to a multi-task model capable of performing at least a first task and a second task using at least one shared. The training module further causes the processor to determine a first value (gradience or loss) for the first task and a second value (gradience or loss) for a second task using the simulation training data and the at least one shared parameter, determine a task induced variance between the first value and the second value, and iteratively adjust the at least one shared parameter to reduce the task induced variance.