公开(公告)号：US11188831B2

公开(公告)日：2021-11-30

申请号：US15796410

申请日：2017-10-27

Applicant: Amazon Technologies, Inc.

Inventor： Benjamin Lev Snyder ,  Liliane Jeanne Barbour ,  Aritra Biswas ,  Simone Elviretti ,  Rasika Sanjay Jangle ,  Paul Hercules Mandac Rivera ,  James Stevenson ,  Daniel Patrick Weaver

IPC: G06N5/02 ,  G06F16/54 ,  G06Q30/06 ,  G06F16/904 ,  G06F16/9038 ,  G06N3/04

Abstract: In response to a programmatic interaction, respective representations of items of an initial result set are presented to an item consumer. One or more result refinement iterations are then conducted. In a given iteration, one or more feedback indicators with respect to one or more items are identified, a machine learning model is trained using at least the feedback indicators to generate respective result set candidacy metrics for at least some items, and the metrics are then used to transmit additional items for presentation to the item consumer.

公开(公告)号：US11068715B1

公开(公告)日：2021-07-20

申请号：US16294284

申请日：2019-03-06

Applicant: Amazon Technologies, Inc.

Inventor： Zhen Zuo ,  Jinmiao Fu ,  Vikash Pansari ,  Yi Sun ,  Phani Harish Wajjala ,  Dheepikaa Natarajan ,  Xiangdong Qian ,  Yikai Ni ,  Jianfeng Lin ,  Nabajyoti Patowary ,  Gaurav Dasgupta ,  James Stevenson

IPC: G10L15/22 ,  G06K9/00 ,  G06Q30/06 ,  G06K9/32

Abstract: Systems, methods, and computer-readable media are disclosed for systems and methods for automated detection and extraction of nutrition information for food products. Example methods may include identifying at least a first image comprising nutritional data, generating a second image including a first portion of the first image, and generating a third image including a second portion of the first image. Some embodiments may include generating a fourth image using the second image, generating a fifth image using the third image, determining, using at least one of the second image and the fourth image, the first text, determining, using at least one of the third image and the fifth image, the second text, and causing at least a portion of the first text and a portion of the second text to be stored in association with a product identifier.

公开(公告)号：US20190130285A1

公开(公告)日：2019-05-02

申请号：US15796410

申请日：2017-10-27

Applicant: Amazon Technologies, Inc.

Inventor： Benjamin Lev Snyder ,  Liliane Jeanne Barbour ,  Aritra Biswas ,  Simone Elviretti ,  Rasika Sanjay Jangle ,  Paul Hercules Mandac Rivera ,  James Stevenson ,  Daniel Patrick Weaver

IPC: G06N5/02 ,  G06N3/04 ,  G06F17/30

Abstract: In response to a programmatic interaction, respective representations of items of an initial result set are presented to an item consumer. One or more result refinement iterations are then conducted. In a given iteration, one or more feedback indicators with respect to one or more items are identified, a machine learning model is trained using at least the feedback indicators to generate respective result set candidacy metrics for at least some items, and the metrics are then used to transmit additional items for presentation to the item consumer.