The patent badge is an abbreviated version of the USPTO patent document. The patent badge does contain a link to the full patent document.
The patent badge is an abbreviated version of the USPTO patent document. The patent badge covers the following: Patent number, Date patent was issued, Date patent was filed, Title of the patent, Applicant, Inventor, Assignee, Attorney firm, Primary examiner, Assistant examiner, CPCs, and Abstract. The patent badge does contain a link to the full patent document (in Adobe Acrobat format, aka pdf). To download or print any patent click here.
Patent No.:
Date of Patent:
Sep. 15, 2026
Filed:
Dec. 04, 2023
Amazon Technologies, Inc., Seattle, WA (US);
Amit Adam, Haifa, IL;
Avi Avraham Ben-Cohen, Holon, IL;
Ran Schley, Kiryat Bialik, IL;
Matan Karklinsky, Zur Hadassa, IL;
Elizabeth Riley, Baltimore, MD (US);
Marc Jason Stromberg, Frisco, TX (US);
Alex Strand, Los Angeles, CA (US);
Scott Patrick Swim, McCordsville, IN (US);
Kevin Soldani, Norwalk, CT (US);
Ianir Ideses, Raanana, IL;
Amazon Technologies, Inc., Seattle, WA (US);
Abstract
Techniques for providing graphic overlays on live videos using a machine learning (ML) model are described. According to some examples, a computer-implemented method includes receiving, at an on-site production service, a live stream of an event; generating, by a machine learning model of the on-site production service, a first inference of a first set of surface-anchored coordinates for a plurality of points of a playing surface of the event in a first frame of the live stream, and a first set of visibility probabilities for the plurality of points; generating, by the machine learning model of the on-site production service, a second inference of a second set of surface-anchored coordinates for at least one of the plurality of points of the playing surface of the event in a second frame of the live stream, and a second visibility probability for the at least one of the plurality of points; and inserting, by the on-site production service, a graphic into the first frame and the second frame of the live stream based at least in part on the first inference of the first set of surface-anchored coordinates for the plurality of points of the playing surface of the event in the first frame, and the second inference of the second set of surface-anchored coordinates for the at least one of the plurality of points of the playing surface of the event in the second frame.