知搜-全球专利检索

  1. 登录
  2. 注册

搜索结果

13 条记录 (耗时 0.012 秒)

    UNSUPERVISED METADATA GENERATION FOR VEHICLE DATA LOGS
    1.
    发明公开
    UNSUPERVISED METADATA GENERATION FOR VEHICLE DATA LOGS 审中-公开

    公开(公告)号：US20230408294A1

    公开(公告)日：2023-12-21

    申请号：US18338834

    申请日：2023-06-21

    申请人： Atieva, Inc.

    发明人： Michael Anderson

    IPC分类号： G01C21/00 ,  G06F16/29 ,  G07C5/00

    CPC分类号： G01C21/3856 ,  G06F16/29 ,  G07C5/008

    摘要： A method of performing unsupervised metadata generation for vehicle data comprises: receiving vehicle data collected during travel by a vehicle, the vehicle data including position data, speed data, and timestamps of the position data and the speed data; defining, using the vehicle data, a map route corresponding to the travel in map data; determining metadata for the travel using the map route; and annotating the vehicle data with the determined metadata.

    SYSTEM AND METHOD OF DATA MODELING, ESTIMATION, AND SELECTIVE CORRECTION FOR AGRICULTURAL MAPS
    2.
    发明公开
    SYSTEM AND METHOD OF DATA MODELING, ESTIMATION, AND SELECTIVE CORRECTION FOR AGRICULTURAL MAPS 审中-公开

    公开(公告)号：US20240255306A1

    公开(公告)日：2024-08-01

    申请号：US18160106

    申请日：2023-01-26

    申请人： Deere & Company

    发明人： Eric Edstrom ,  Noel W. Anderson

    IPC分类号： G01C21/00

    CPC分类号： G01C21/3837 ,  G01C21/3856

    摘要： A system and method are provided for generating maps based on map types further associated with various data entry fields needed to populate the respective map. A first set of data entry fields are populated for which underlying information is available from at least one data source, and a second set is identified of any data entry fields for which at least some underlying information is missing after the first set is populated. The method further includes selecting estimated information to populate the second set of data entry fields, generating the map including the underlying information and the estimated information in the corresponding data entry fields, and displaying the map on a user interface, wherein the displayed map includes an error estimate indication corresponding to at least one of the second set of data entry fields on at least a portion thereof.

    HIGH-DEFINITION ENERGY CONSUMPTION MAP FOR VEHICLES
    4.
    发明公开
    HIGH-DEFINITION ENERGY CONSUMPTION MAP FOR VEHICLES 审中-公开

    公开(公告)号：US20240118098A1

    公开(公告)日：2024-04-11

    申请号：US17961450

    申请日：2022-10-06

    申请人： Rivian IP Holdings, LLC

    发明人： Brennan Matthew BOBLETT ,  Jason Meyer QUINT

    IPC分类号： G01C21/34 ,  G01C21/00 ,  G01C21/36

    CPC分类号： G01C21/3469 ,  G01C21/3415 ,  G01C21/3617 ,  G01C21/367 ,  G01C21/3856 ,  G01C21/3896

    摘要： An in-vehicle control system for providing an energy consumption map for a vehicle is provided. The method includes displaying, on a display of a computing system of the vehicle, a map of an environment external to the vehicle. The map includes a predicted drive path extending between an initial location of the vehicle and a final destination of the vehicle within the environment. The method further includes generating, by the computing system, and based on the predicted drive path, a prediction of the energy consumption of the vehicle along the one or more trajectories of the predicted drive path, and displaying, on the display of the computing system, the predicted energy consumption of the vehicle along the one or more trajectories of the predicted drive path. The predicted energy consumption of the vehicle is displayed so as to at least partially overlay the predicted drive path.

    TRAINING MACHINE LEARNING MODELS BASED ON MOVEMENT AND TIMING DATA
    10.
    发明公开
    TRAINING MACHINE LEARNING MODELS BASED ON MOVEMENT AND TIMING DATA 审中-公开

    公开(公告)号：US20240159555A1

    公开(公告)日：2024-05-16

    申请号：US18055665

    申请日：2022-11-15

    申请人： Capital One Services, LLC

    发明人： Matthew Louis NOWAK ,  Michael Anthony YOUNG, JR. ,  Christopher McDANIEL

    IPC分类号： G01C21/00 ,  G06N20/00

    CPC分类号： G01C21/383 ,  G01C21/3856 ,  G06N20/00

    摘要： Systems and methods are disclosed herein for training machine learning models using precision data based on movement and timing of data collection. The system trains the machine learning model to predict object locations in an environment. The training data includes location data, timing data, and motion data, collection of which is triggered by a user input indicating that a particular object was located. The system generates precision parameters for the locations by assigning initial values to entries within the training data and may lower precision parameters based on the timing data indicating that an object was located faster than a threshold time or based on the motion data indicating that a rate of motion exceeded a threshold rate when a corresponding object was located. The system may update the training data with the lowered precision parameters and train the machine learning model to predict the object locations within the environment.