-
公开(公告)号：US11151628B2
公开(公告)日：2021-10-19
申请号：US16290607
申请日：2019-03-01
Applicant: Capital One Services, LLC
Inventor： Stephen Wylie , Staevan Duckworth , Qiaochu Tang , Geoffrey Dagley , Olalekan Awoyemi
IPC: G06Q30/00 , G06Q30/06 , H04W4/02 , G06F3/0482 , G06F3/0484 , G06K9/00
Abstract: A device may determine that a user device, associated with a user, and a first object for comparison are within a threshold proximity of one another. The device may communicate with the user device to obtain a comparison profile, associated with the user, based on determining that the user device and the first object are within the threshold proximity of one another. The device may identify, based on the comparison profile, a set of desired attributes. The device may identify, based on the comparison profile, a set of attribute categories to be used to compare the first object and the set of desired attributes. The device may determine a first set of attributes of the first object for the set of attribute categories. The device may present, for display, information that identifies one or more attributes of the first set of attributes or the set of desired attributes.
-
公开(公告)号：US10949672B1
公开(公告)日：2021-03-16
申请号：US16832174
申请日：2020-03-27
Applicant: Capital One Services, LLC
Inventor： Qiaochu Tang , Micah Price , Jason Hoover , Geoffrey Dagley , Stephen Wylie
IPC: G06K9/00 , G06N3/08 , G06N3/04 , G06T7/70 , G06T7/00 , G09B5/02 , G09B19/00 , G06T7/20 , G06T11/00
Abstract: A device having extended reality capabilities may capture a video feed including one or more video frames depicting an object that is visible in a field of view of the device. The device may provide the one or more video frames to a machine learning model that identifies the object and one or more parts of the object that are depicted in the one or more video frames. The device may obtain positional tracking information that represents a position and an orientation associated with the object relative to a coordinate space that corresponds to the field of view of the device. The device may obtain a workflow including a sequence of content items for visually inspecting the object using the extended reality capabilities of the device. The device may render digital content associated with the workflow using the extended reality capabilities of the device.
-
公开(公告)号：US20200349214A1
公开(公告)日：2020-11-05
申请号：US16597839
申请日：2019-10-09
Applicant: Capital One Services, LLC
Inventor： Qiaochu TANG , Stephen Wylie , Jason Hoover , Avid Ghamsari , Micah Price , Geoffrey Dagley
IPC: G06F16/9538 , G06N20/00 , G06F17/27
Abstract: Techniques to leverage machine learning for search engine optimization (SEO). Some techniques are applicable to turnkey e-commerce solutions operating on the Internet as an online platform. These techniques use machine learning to gain insights into patterns/relationships between search terms and items for sale on the online platform. Depending on a relevancy or an accuracy of a search term to a particular item for sale, the online platform inserts the search term into the online platform such that a public search engine crawler increases a search engine result position for the online platform. Other embodiments are described and claimed.
-
74.发明授权
公开(公告)号：US10803544B2
公开(公告)日：2020-10-13
申请号：US16425279
申请日：2019-05-29
Applicant: Capital One Services, LLC
Inventor： Qiaochu Tang , Geoffrey Dagley , Micah Price , Sunil Vasisht , Stephen Wylie , Jason Hoover
Abstract: The disclosed technology includes systems and methods for enhancing machine vision object recognition based on a plurality of captured images and an accumulation of corresponding classification analysis scores. A method is provided for capturing, with a camera of a mobile computing device, a plurality of images, each image of the plurality of images comprising a first object. The method includes processing, with a classification module comprising a trained neural network processing engine, at least a portion of the plurality of images. The method includes generating, with the classification module and based on the processing, one or more object classification scores associated with the first object. The method includes accumulating, with an accumulating module, the one or more object classification scores. And responsive to a timeout or an accumulated score exceeding a predetermined threshold, the method includes outputting classification information of the first object.
-
公开(公告)号：US20200302381A1
公开(公告)日：2020-09-24
申请号：US16894979
申请日：2020-06-08
Applicant: Capital One Services, LLC
Inventor： Jason Hoover , Qiaochu Tang , Micah Price , Geoffrey Dagley , Stephen Wylie
IPC: G06Q10/08 , G06K9/62 , G06F16/245
Abstract: In one aspect, the present disclosure relates to a method for improved efficiency of an inventory management system. The method can include: performing a first inventory search of inventory data stored in a database using search criteria; receiving, from a beacon sensor, inventory data broadcast from a plurality of beacons; performing a second inventory search by correlating the inventory data broadcast from the plurality of beacons with results of the first inventory search; receiving image data from a camera; performing image classification on the image data to determine one or more class labels for an inventory item within the field-of-view of the camera; performing a third inventory search by correlating the class labels with results of the second inventory search; and presenting results of the third inventory search to a user.
-
公开(公告)号：US20200293821A1
公开(公告)日：2020-09-17
申请号：US16882962
申请日：2020-05-26
Applicant: Capital One Services, LLC
Inventor： Micah Price , Jason Hoover , Geoffrey Dagley , Stephen Wylie , Qiaochu Tang
Abstract: In one aspect, the present disclosure relates to a method for or performing single-pass object detection and image classification. The method comprises receiving image data for an image in a system comprising a convolutional neural network (CNN), the CNN comprising a first convolutional layer, a last convolutional layer, and a fully connected layer; providing the image data to an input of the first convolutional layer; extracting multi-channel data from the output of the last convolutional layer; and summing the extracted data to generate a general activation map; and detecting a location of an object within the image by applying the general activation map to the image data.
-
公开(公告)号：US20200216026A1
公开(公告)日：2020-07-09
申请号：US16243499
申请日：2019-01-09
Applicant: Capital One Services, LLC
Inventor： Micah PRICE , Qiaochu Tang , Stephen Wylie
IPC: B60R25/30 , G08B13/196 , B60R25/102 , B60R25/33
Abstract: A server device obtains, from a primary device associated with a first vehicle, event data concerning an event, video data concerning the event, and/or location data concerning a location of the first vehicle. The server device processes the event data to determine a type of the event, and determines, based on the type of the event and the location data, a proximity zone around the location of the first vehicle. The server device determines additional devices within the proximity zone, where each additional device is associated with a different vehicle. The server device sends, to the additional devices, a message requesting additional video data concerning the event, and obtains the additional video data from at least one additional device. The server device performs, based on relevant event information based on the event data, the video data, and the additional video data, an action concerning the event.
-
公开(公告)号：US10671886B2
公开(公告)日：2020-06-02
申请号：US16253490
申请日：2019-01-22
Applicant: Capital One Services, LLC
Inventor： Micah Price , Jason Hoover , Geoffrey Dagley , Stephen Wylie , Qiaochu Tang
Abstract: In one aspect, the present disclosure relates to a method for or performing single-pass object detection and image classification. The method comprises receiving image data for an image in a system comprising a convolutional neural network (CNN), the CNN comprising a first convolutional layer, a last convolutional layer, and a fully connected layer; providing the image data to an input of the first convolutional layer; extracting multi-channel data from the output of the last convolutional layer; and summing the extracted data to generate a general activation map; and detecting a location of an object within the image by applying the general activation map to the image data.
-
公开(公告)号：US10593128B1
公开(公告)日：2020-03-17
申请号：US16546286
申请日：2019-08-20
Applicant: Capital One Services, LLC
Inventor： Micah Price , Geoffrey Dagley , Qiaochu Tang , Jason Hoover , Stephen Wylie
Abstract: Techniques for providing indirect local geo-positioning using AR markers are disclosed. A first moveable AR marker can be located or found by a computing device. A location of the first moveable AR marker can be known and shared with the computing device. The location of the first moveable AR marker can be based on distance between the first moveable AR marker and a fixed AR marker. A distance to the first moveable AR marker can be determined. Based on the known location of the first moveable AR marker and the distance to the first moveable AR marker from the computing device, an estimate of the location of the computing device can be determined without having line-of-sight (LOS) to the fixed AR marker.
-
80.发明申请
公开(公告)号：US20190279329A1
公开(公告)日：2019-09-12
申请号：US16425279
申请日：2019-05-29
Applicant: Capital One Services, LLC
Inventor： Qiaochu Tang , Geoffrey Dagley , Micah Price , Sunil Vasisht , Stephen Wylie , Jason Hoover
Abstract: The disclosed technology includes systems and methods for enhancing machine vision object recognition based on a plurality of captured images and an accumulation of corresponding classification analysis scores. A method is provided for capturing, with a camera of a mobile computing device, a plurality of images, each image of the plurality of images comprising a first object. The method includes processing, with a classification module comprising a trained neural network processing engine, at least a portion of the plurality of images. The method includes generating, with the classification module and based on the processing, one or more object classification scores associated with the first object. The method includes accumulating, with an accumulating module, the one or more object classification scores. And responsive to a timeout or an accumulated score exceeding a predetermined threshold, the method includes outputting classification information of the first object.
