US 12,267,305 B2
Privacy preserving document analysis
Nikolaos Barmpalios, Sunnyvale, CA (US); Ruchi Rajiv Deshpande, Belmont, CA (US); Randy Lee Swineford, Sunnyvale, CA (US); Nargol Rezvani, Sunnyvale, CA (US); Andrew Marc Greene, Newton, MA (US); Shawn Alan Gaither, Raleigh, NC (US); and Michael Kraley, Lexington, MA (US)
Assigned to Adobe Inc., San Jose, CA (US)
Filed by Adobe Inc., San Jose, CA (US)
Filed on May 15, 2023, as Appl. No. 18/317,338.
Application 18/317,338 is a continuation of application No. 16/695,636, filed on Nov. 26, 2019, granted, now 11,689,507.
Prior Publication US 2023/0336532 A1, Oct. 19, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. H04L 9/40 (2022.01); G06N 5/04 (2023.01); G06N 20/00 (2019.01); G06Q 30/0202 (2023.01)
CPC H04L 63/04 (2013.01) [G06N 5/04 (2013.01); G06N 20/00 (2019.01); G06Q 30/0202 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
receiving, by a processing device, a plurality of stamp representations each corresponding to a respective source document and representing visual features of content of the respective source document without containing the content of the respective source document;
generating, by the processing device and using a machine learning model, a stamp embedding space by processing the plurality of stamp representations;
receiving, by the processing device, an additional stamp representation corresponding to an additional document and having features identifying a type of the additional document; and
identifying, by the processing device, an insight pertaining to the additional document, including the type of the additional document, by comparing the additional stamp representation to the stamp embedding space.