公开(公告)号：US20240427756A1

公开(公告)日：2024-12-26

申请号：US18786068

申请日：2024-07-26

Applicant: Lyft, Inc.

Inventor： Jason Noah Laska ,  Han Suk Kim ,  Matthew Jonathan Feldman ,  Artur Dryomov ,  Lev Dragunov ,  Yakau Bubnou ,  Yanina Michukova ,  Oleg Shnitko ,  Stanislau Binko ,  Siddharth Vijayakrishnan ,  Mark Edward Huberty ,  Kathryn Flaherty Frisbie ,  Mikhail Viktorovich Chuikin ,  Dzmitry Kavaliou

IPC: G06F16/23 ,  G01C21/00 ,  G06F16/29 ,  G06V20/54 ,  H04N7/18

Abstract: Examples disclosed herein involve a computing system configured to (i) receive image data captured by an image-capture device; (ii) based on the received image data, generate a set of map update tasks that each define a respective activity for evaluating whether to update a map; (iii) use a multi-factor prioritization scheme to prioritize the set of map update tasks; (iv) assign at least a subset of map update tasks from the set of map update tasks to one or more curators in accordance with the prioritization of the set of map update tasks; (v) receive, from a client station associated with a given curator via a network-based communication path, data defining feedback from the given curator regarding a given map update task; and (vi) update the map based on the received feedback.

公开(公告)号：US20210271876A1

公开(公告)日：2021-09-02

申请号：US17241791

申请日：2021-04-27

Applicant: Lyft, Inc.

Inventor： Deeksha Goyal ,  Han Suk Kim ,  James Kevin Murphy ,  Albert Yuen

IPC: G06K9/00 ,  G06K9/62 ,  G06K9/46

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media for identifying traffic control features based on telemetry patterns within digital image representations of vehicle telemetry information. The disclosed systems can generate a digital image representation based on collected telemetry information to represent the frequency of different speed-location combinations for transportation vehicles passing through a traffic area. The disclosed systems can also apply a convolutional neural network to analyze the digital image representation and generate a predicted classification of a type of traffic control feature that corresponds to the digital image representation of vehicle telemetry information. The disclosed systems further train the convolutional neural network to determine traffic control features based on training data.

公开(公告)号：US20230205753A1

公开(公告)日：2023-06-29

申请号：US17564686

申请日：2021-12-29

Applicant: Lyft, Inc.

Inventor： Jason Noah Laska ,  Han Suk Kim ,  Matthew Jonathan Feldman ,  Artur Dryomov ,  Lev Dragunov ,  Yakau Bubnou ,  Yanina Michukova ,  Oleg Shnitko ,  Stanislau Binko ,  Siddharth Vijayakrishnan ,  Mark Edward Huberty ,  Kathryn Flaherty Frisbie ,  Mikhail Viktorovich Chuikin ,  Dzmitry Kavaliou

IPC: G06F16/23 ,  G06V20/54 ,  G06F16/29 ,  H04N7/18 ,  G01C21/00

CPC classification number: G06F16/2365 ,  G06V20/54 ,  G06F16/29 ,  H04N7/183 ,  G01C21/3837 ,  G06V2201/10

Abstract: Examples disclosed herein involve a computing system configured to (i) receive, from an image-capture device, image-capture metadata that provides information about a set of images that were passively captured by the image-capture device, (ii) based at least on the received image-capture metadata, apply a collection policy for determining whether to collect any of the set of images that have been passively captured by the image-capture device and thereby determine that a selected subset of one or more images in the set of images are to be collected from the image-capture device, (iii) instruct the image-capture device to upload the selected subset of one or more images, and (iv) after instructing the image-capture device to upload the selected subset of one or more images, receive the selected subset of one or more images from the image-capture device.

公开(公告)号：US10990819B2

公开(公告)日：2021-04-27

申请号：US16408168

申请日：2019-05-09

Applicant: Lyft, Inc.

Inventor： Deeksha Goyal ,  Han Suk Kim ,  James Kevin Murphy ,  Albert Yuen

IPC: G06K9/00 ,  G06K9/62 ,  G06K9/46

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media for identifying traffic control features based on telemetry patterns within digital image representations of vehicle telemetry information. The disclosed systems can generate a digital image representation based on collected telemetry information to represent the frequency of different speed-location combinations for transportation vehicles passing through a traffic area. The disclosed systems can also apply a convolutional neural network to analyze the digital image representation and generate a predicted classification of a type of traffic control feature that corresponds to the digital image representation of vehicle telemetry information. The disclosed systems further train the convolutional neural network to determine traffic control features based on training data.

公开(公告)号：US12050589B2

公开(公告)日：2024-07-30

申请号：US17564686

申请日：2021-12-29

Applicant: Lyft, Inc.

Inventor： Jason Noah Laska ,  Han Suk Kim ,  Matthew Jonathan Feldman ,  Artur Dryomov ,  Lev Dragunov ,  Yakau Bubnou ,  Yanina Michukova ,  Oleg Shnitko ,  Stanislau Binko ,  Siddharth Vijayakrishnan ,  Mark Edward Huberty ,  Kathryn Flaherty Frisbie ,  Mikhail Viktorovich Chuikin ,  Dzmitry Kavaliou

IPC: G06F16/23 ,  G01C21/00 ,  G06F16/29 ,  G06V20/54 ,  H04N7/18

CPC classification number: G06F16/2365 ,  G01C21/3837 ,  G06F16/29 ,  G06V20/54 ,  H04N7/183 ,  G06V2201/10

Abstract: Examples disclosed herein involve a computing system configured to (i) receive, from an image-capture device, image-capture metadata that provides information about a set of images that were passively captured by the image-capture device, (ii) based at least on the received image-capture metadata, apply a collection policy for determining whether to collect any of the set of images that have been passively captured by the image-capture device and thereby determine that a selected subset of one or more images in the set of images are to be collected from the image-capture device, (iii) instruct the image-capture device to upload the selected subset of one or more images, and (iv) after instructing the image-capture device to upload the selected subset of one or more images, receive the selected subset of one or more images from the image-capture device.

公开(公告)号：US11694426B2

公开(公告)日：2023-07-04

申请号：US17241791

申请日：2021-04-27

Applicant: Lyft, Inc.

Inventor： Deeksha Goyal ,  Han Suk Kim ,  James Kevin Murphy ,  Albert Yuen

IPC: G06K9/00 ,  G06K9/62 ,  G06K9/46 ,  G06V10/50 ,  G06F18/21 ,  G06F18/2431 ,  G06V10/764 ,  G06V10/82 ,  G06V20/58

CPC classification number: G06V10/50 ,  G06F18/217 ,  G06F18/2431 ,  G06V10/764 ,  G06V10/82 ,  G06V20/582

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media for identifying traffic control features based on telemetry patterns within digital image representations of vehicle telemetry information. The disclosed systems can generate a digital image representation based on collected telemetry information to represent the frequency of different speed-location combinations for transportation vehicles passing through a traffic area. The disclosed systems can also apply a convolutional neural network to analyze the digital image representation and generate a predicted classification of a type of traffic control feature that corresponds to the digital image representation of vehicle telemetry information. The disclosed systems further train the convolutional neural network to determine traffic control features based on training data.