公开(公告)号：US11944215B2

公开(公告)日：2024-04-02

申请号：US17699484

申请日：2022-03-21

Applicant: NCR Voyix Corporation

Inventor： Qian Yang ,  Frank Douglas Hinek ,  Dharamendra Kumar ,  Matias Gabriel Szylkowski Resnikow ,  Brent Vance Zucker

IPC: G06Q20/00 ,  A47F9/04 ,  G06F3/0488 ,  G06N20/00 ,  G06Q20/18 ,  G06Q20/20 ,  G06T7/73 ,  G06V10/40 ,  G06V10/764 ,  G06V10/80 ,  G06V20/64 ,  G07G1/00 ,  G06V20/68

CPC classification number: A47F9/048 ,  G06F3/0488 ,  G06N20/00 ,  G06Q20/18 ,  G06Q20/208 ,  G06T7/74 ,  G06V10/40 ,  G06V10/764 ,  G06V10/809 ,  G06V20/64 ,  G07G1/0063 ,  G07G1/0072 ,  G06T2207/20132 ,  G06V20/68 ,  G06V2201/10

Abstract: A vison-based self-checkout terminal is provided. Purchased items are placed on a base and multiple cameras take multiple images of each item placed on the base. A location for each item placed on the base is determined along with a depth and the dimensions of each item at its given location on the base. Each item's images are then cropped, and item recognition is performed for each item on that item's cropped images with that item's corresponding depth and dimension attributes. An item identifier for each item is obtained along with a corresponding price and a transaction associated with items are completed.

公开(公告)号：US20240324797A1

公开(公告)日：2024-10-03

申请号：US18623799

申请日：2024-04-01

Applicant: NCR Voyix Corporation

Inventor： Qian Yang ,  Frank Douglas Hinek ,  Dharamendra Kumar ,  Matias Gabriel Szylkowski Resnikow ,  Brent Vance Zucker

IPC: A47F9/04 ,  G06F3/0488 ,  G06N20/00 ,  G06Q20/18 ,  G06Q20/20 ,  G06T7/73 ,  G06V10/40 ,  G06V10/764 ,  G06V10/80 ,  G06V20/64 ,  G06V20/68 ,  G07G1/00

CPC classification number: A47F9/048 ,  G06F3/0488 ,  G06N20/00 ,  G06Q20/18 ,  G06Q20/208 ,  G06T7/74 ,  G06V10/40 ,  G06V10/764 ,  G06V10/809 ,  G06V20/64 ,  G07G1/0063 ,  G07G1/0072 ,  G06T2207/20132 ,  G06V20/68 ,  G06V2201/10

Abstract: A vision-based self-checkout terminal is provided. Purchased items are placed on a base and multiple cameras take multiple images of each item placed on the base. A location for each item placed on the base is determined along with a depth and the dimensions of each item at its given location on the base. Each item's images are then cropped, and item recognition is performed for each item on that item's cropped images with that item's corresponding depth and dimension attributes. An item identifier for each item is obtained along with a corresponding price and a transaction associated with items are completed.

公开(公告)号：US20240289643A1

公开(公告)日：2024-08-29

申请号：US18115624

申请日：2023-02-28

Applicant: NCR Voyix Corporation

Inventor： Pragya Tripathi ,  Yanxin Ye ,  Brij Hareshbhai Rokad ,  Dharamendra Kumar

IPC: G06N5/022

CPC classification number: G06N5/022

Abstract: An enterprise's data source relevant to their customers is obtained at predefined intervals of time. The data is processed through classification machine learning models (MLMs) and labeled with features. The labeled data is provided as input to an attrition predicting MLM and one or more lists are provided as output identifying customers likely to leave the enterprise and customers with a high likelihood of remaining with the enterprise when provided an incentive to do so. The one or more lists are provided to enterprise interfaces and/or promotion systems for mitigating customer attrition. In an embodiment, results for the attrition predicting MLM are compared against results predicted by a Recency, Frequency, Monetary (RFM) analyzer in view of subsequent actual observed results for the customers with the enterprise. A continuous feedback loop for retaining the attrition prediction MLM is processed based on the comparison to improve the prediction MLM's F1 accuracy metric.