US 12,067,573 B2
Method, computer, and program for artwork management
Lachezar Sashov Dodov, Sofia (BG); and Andreas Ludwig, Dusseldorf (DE)
Assigned to Wacom Co., Ltd., Saitama (JP)
Filed by Wacom Co., Ltd., Saitama (JP)
Filed on Oct. 27, 2022, as Appl. No. 17/975,072.
Application 17/975,072 is a continuation of application No. PCT/JP2021/017839, filed on May 11, 2021.
Claims priority of application No. 2020-088630 (JP), filed on May 21, 2020.
Prior Publication US 2023/0044309 A1, Feb. 9, 2023
Int. Cl. G06Q 20/12 (2012.01); G06Q 20/38 (2012.01); G06Q 20/40 (2012.01); G06Q 50/18 (2012.01)
CPC G06Q 20/4016 (2013.01) [G06Q 20/389 (2013.01); G06Q 50/184 (2013.01)] 12 Claims
OG exemplary drawing
 
1. A server used in a system for monitoring unauthorized use of digital artwork displayed on a website, the server configured to:
receive, from a terminal, an authenticity determination request regarding a target artwork displayed on a website, wherein the authenticity determination request includes artist information regarding an artist associated with the target artwork;
input one or more values included in stroke data associated with the target artwork to a machine learning model, wherein the machine learning model is generated based on training an artificial intelligence (AI) program with the one or more values which indicate features of artwork of the artist and which are included in stroke data of at least one reference artwork associated with the same artist;
wherein the one or more values are indicative of at least one of an average value and a dispersion of brush stroke speed, an average value and a dispersion of a pen pressure value, an average value and a dispersion of pen angle data, or time allocation of a pen touch state and a pen hover state;
acquire a target artist feature value associated with the target artwork, which is output from the machine learning model;
determine authenticity of the target artwork by comparing the acquired target artist feature value and a reference artist feature value previously determined by the machine learning model that was trained with the stroke data of at least one reference artwork associated with the artist; and
provide a result of the determined authenticity of the target artwork to the terminal.