Neural networks for object detection

    公开(公告)号:US10013773B1

    公开(公告)日:2018-07-03

    申请号:US15381288

    申请日:2016-12-16

    Applicant: Waymo LLC

    Abstract: A neural network system for identifying positions of objects in an input image can include an object detector neural network, a memory interface subsystem, and an external memory. The object detector neural network is configured to, at each time step of multiple successive time steps, (i) receive a first neural network input that represents the input image and a second neural network input that identifies a first set of positions of the input image that have each been classified as showing a respective object of the set of objects, and (ii) process the first and second inputs to generate a set of output scores that each represents a respective likelihood that an object that is not one of the objects shown at any of the positions in the first set of positions is shown at a respective position of the input image that corresponds to the output score.

    Construction object detection
    14.
    发明授权

    公开(公告)号:US09767370B1

    公开(公告)日:2017-09-19

    申请号:US15215890

    申请日:2016-07-21

    Applicant: Waymo LLC

    Abstract: Aspects of the disclosure relate to identifying construction objects. As an example, an image captured by a camera associated with a vehicle as the vehicle is driven along a roadway may be received. This image may be converted into a first channel corresponding to an average brightness contribution from red, blue and green channels of the image. The image may also be converted into a second channel corresponding to a contribution of a color from the red and the green channels of the image. A template may then be used to identify a region of the image corresponding to a potential construction object from the first channel and the second channel.

    Vehicle control using vision-based flashing light signal detection

    公开(公告)号:US11281230B2

    公开(公告)日:2022-03-22

    申请号:US16837017

    申请日:2020-04-01

    Applicant: Waymo LLC

    Abstract: A computing device of a first vehicle may receive a first image and a second image of a second vehicle having flashing light signals. The computing device may determine, in the first image and the second image, an image region that bounds the second vehicle such that the image region substantially encompasses the second vehicle. The computing device may determine a polar grid that partitions the image region in the first image and the second image into polar bins, and identify portions of image data exhibiting a change in color and a change in brightness between the first image and the second image. The computing device may determine a type of the flashing light signals and a type of the second vehicle; and accordingly provide instructions to control the first vehicle.

    OBJECT LOCALIZATION USING MACHINE LEARNING

    公开(公告)号:US20210364637A1

    公开(公告)日:2021-11-25

    申请号:US17394861

    申请日:2021-08-05

    Applicant: Waymo LLC

    Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for determining a location of a particular object relative to a vehicle. In one aspect, a method includes obtaining sensor data captured by one or more sensors of a vehicle. The sensor data is processed by a convolutional neural network to generate a sensor feature representation of the sensor data. Data is obtained which defines a particular spatial region in the sensor data that has been classified as including sensor data that characterizes the particular object. An object feature representation of the particular object is generated from a portion of the sensor feature representation corresponding to the particular spatial region. The object feature representation of the particular object is processed using a localization neural network to generate the location of the particular object relative to the vehicle.

    Object localization using machine learning

    公开(公告)号:US11105924B2

    公开(公告)日:2021-08-31

    申请号:US16151880

    申请日:2018-10-04

    Applicant: Waymo LLC

    Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for determining a location of a particular object relative to a vehicle. In one aspect, a method includes obtaining sensor data captured by one or more sensors of a vehicle. The sensor data is processed by a convolutional neural network to generate a sensor feature representation of the sensor data. Data is obtained which defines a particular spatial region in the sensor data that has been classified as including sensor data that characterizes the particular object. An object feature representation of the particular object is generated from a portion of the sensor feature representation corresponding to the particular spatial region. The object feature representation of the particular object is processed using a localization neural network to generate the location of the particular object relative to the vehicle.

    Assisted Perception For Autonomous Vehicles

    公开(公告)号:US20210191420A1

    公开(公告)日:2021-06-24

    申请号:US17192089

    申请日:2021-03-04

    Applicant: Waymo LLC

    Abstract: Disclosed herein are systems and methods for providing supplemental identification abilities to an autonomous vehicle system. The sensor unit of the vehicle may be configured to receive data indicating an environment of the vehicle, while the control system may be configured to operate the vehicle. The vehicle may also include a processing unit configured to analyze the data indicating the environment to determine at least one object having a detection confidence below a threshold. Based on the at least one object having a detection confidence below a threshold, the processor may communicate at least a subset of the data indicating the environment for further processing. The vehicle is also configured to receive an indication of an object confirmation of the subset of the data. Based on the object confirmation of the subset of the data, the processor may alter the control of the vehicle by the control system.

    Stop sign detection and response
    20.
    发明授权

    公开(公告)号:US10401859B1

    公开(公告)日:2019-09-03

    申请号:US15915383

    申请日:2018-03-08

    Applicant: Waymo LLC

    Abstract: Aspects of the disclosure relate to detecting and responding to stop signs. An object detected in a vehicle's environment having location coordinates may be identified as a stop sign and, it may be determined whether the location coordinates of the identified stop sign correspond to a location of a stop sign in detailed map information. Then, whether the identified stop sign applies to the vehicle may be determined based on the detailed map information or on a number of factors. Then, if the identified stop sign is determined to apply to the vehicle, responses of the vehicle to the stop sign may be determined, and, the vehicle may be controlled based on the determined responses.

Patent Agency Ranking