利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

10 条记录 (耗时 0.017 秒)

    Method for Temporal Correction of Multimodal Data
    3.
    发明公开
    Method for Temporal Correction of Multimodal Data 审中-公开

    公开(公告)号：US20230419649A1

    公开(公告)日：2023-12-28

    申请号：US18337153

    申请日：2023-06-19

    申请人： Robert Bosch GmbH

    发明人： Claudius Glaeser ,  Fabian Timm ,  Florian Drews ,  Michael Ulrich ,  Florian Faion ,  Lars Rosenbaum

    IPC分类号： G06V10/82 ,  G06V10/75 ,  G06V10/80 ,  G06V10/776

    CPC分类号： G06V10/82 ,  G06V10/751 ,  G06V10/80 ,  G06V10/776

    摘要： A method for the chronological correction of multimodal data includes:



    receiving a first data set from a reference sensor with measurements at different measurement timepoints,
    receiving a second data set of a second sensor with measurements at different measurement timepoints, each not exactly matching those of the reference sensor,
    reading the first and the second data sets by a neural network and identifying a respective plurality of feature vectors for the first and second data set at the respective measurement timepoints,
    merging and comparing the respective feature vectors, which refer to corresponding, not exactly matching measurement timepoints, by the neural network so that parameters of a chronological correction are identified, and
    identifying a chronological offset between the respective measurement timepoints of the reference sensor and the second sensor, and/or a corrected data set from the second sensor based on the measurement timepoints of the reference sensor.

    METHOD AND DEVICE FOR COMPRESSING A NEURAL NETWORK
    5.
    发明申请
    METHOD AND DEVICE FOR COMPRESSING A NEURAL NETWORK 有权

    公开(公告)号：US20220076124A1

    公开(公告)日：2022-03-10

    申请号：US17395845

    申请日：2021-08-06

    申请人： Robert Bosch GmbH

    发明人： Fabian Timm ,  Lukas Enderich

    IPC分类号： G06N3/08

    摘要： A method for compressing a neural network. The method includes: defining a maximum complexity of the neural network; ascertaining a first cost function; ascertaining a second cost function, which characterizes a deviation of a current complexity of the neural network in relation to the defined complexity; training the neural network in such a way that a sum of a first and a second cost function is optimized as a function of parameters of the neural network; and removing those weightings whose assigned scaling factor is smaller than a predefined threshold value.

    METHOD AND SYSTEM FOR GENERATING RADAR REFLECTION POINTS
    6.
    发明申请
    METHOD AND SYSTEM FOR GENERATING RADAR REFLECTION POINTS 审中-公开

    公开(公告)号：US20200379087A1

    公开(公告)日：2020-12-03

    申请号：US16878814

    申请日：2020-05-20

    申请人： Robert Bosch GmbH

    发明人： Chun Yang ,  Sebastian Muenzner ,  Zoltan Karasz ,  Fabian Timm ,  Jasmin Ebert ,  Laszlo Anka

    IPC分类号： G01S7/41 ,  G01S13/89

    摘要： A method for generating radar reflection points comprising the steps of: providing a plurality of predefined radar reflection points of at least one first object detected by a radar and at least one first scenario description describing a first environment related to the detected first object; converting the predefined radar reflection points into at least one first power distribution pattern image related to a distribution of a power returning from the detected first object; training a model based on the first power distribution pattern image and the first scenario description; providing at least one second scenario description describing a second environment related to a second object; generating at least one second power distribution pattern image related to a distribution of a power returning from the second object based on the trained model and the second scenario description; and sampling the second power distribution pattern image.

    FAST QUANTISED TRAINING OF TRAINABLE MODULES
    9.
    发明申请
    FAST QUANTISED TRAINING OF TRAINABLE MODULES 有权

    公开(公告)号：US20220277200A1

    公开(公告)日：2022-09-01

    申请号：US17632735

    申请日：2020-08-06

    申请人： Robert Bosch GmbH

    发明人： Fabian Timm ,  Lukas Enderich

    IPC分类号： G06N3/08

    摘要： A method for training a trainable module that maps input variables onto output variables through an internal processing chain. A learning data set is provided including learning values of the input variables and associated learning values of the output variables. A list of discrete values is provided from which the parameters characterizing the internal processing chain are to be selected, the discrete values being selected such that they can be stored without loss of quality. The learning values are mapped by the trainable module onto assessment values of the output variables. A cost function is evaluated that characterizes deviations of the assessment values of the output variables from the learning values and of at least one parameter of the internal processing chain from at least one discrete value in the list. At least one parameter of the internal processing chain is adjusted to improve the value of the cost function.

    Determining a state of the surrounding area of a vehicle, using linked classifiers
    10.
    发明授权
    Determining a state of the surrounding area of a vehicle, using linked classifiers 有权

    公开(公告)号：US11100337B2

    公开(公告)日：2021-08-24

    申请号：US16387292

    申请日：2019-04-17

    申请人： Robert Bosch GmbH

    发明人： Fabian Timm

    IPC分类号： G06K9/00 ,  G06K9/62

    摘要： A method for determining a state of the surrounding area of a vehicle includes: receiving sensor data of at least one surrounding-area sensor of the vehicle; feeding at least a first portion of the sensor data into at least one first classifier; generating an intermediate probability from the first portion of the sensor data, using the first classifier; feeding at least a second portion of the sensor data and the at least one intermediate probability into a second classifier; generating a final probability of the state of the surrounding area from the second portion of the sensor data and the at least one intermediate probability, using the second classifier.