US 11,893,829 B2
Method for deploying a face sample library and method and apparatus for business processing based on face recognition
Xi Shen, Shanghai (CN); Jialiang Kang, Shanghai (CN); and Jien Zhou, Shanghai (CN)
Assigned to CHINA UNION PAY CO., LTD., Shanghai (CN)
Appl. No. 17/268,682
Filed by CHINA UNIONPAY CO., LTD., Shanghai (CN)
PCT Filed Jan. 22, 2020, PCT No. PCT/CN2020/073822
§ 371(c)(1), (2) Date Feb. 16, 2021,
PCT Pub. No. WO2020/220779, PCT Pub. Date Nov. 5, 2020.
Claims priority of application No. 201910362691.2 (CN), filed on Apr. 30, 2019.
Prior Publication US 2021/0192191 A1, Jun. 24, 2021
Int. Cl. G06V 40/16 (2022.01); G06F 16/58 (2019.01); G06Q 10/105 (2023.01); G06V 40/50 (2022.01)
CPC G06V 40/172 (2022.01) [G06F 16/5866 (2019.01); G06Q 10/105 (2013.01); G06V 40/50 (2022.01)] 17 Claims
OG exemplary drawing
 
1. A method for deploying a face sample library, comprising:
obtaining historical business information of users corresponding to face samples in a password face sample library under a condition that a number of face samples in the password face sample library is greater than or equal to a preset sample number threshold, wherein confidential business passwords of the users corresponding to the face samples in the same password face sample library are the same;
dividing the password face sample library to obtain M levels of face sample sub-libraries according to the historical business information, where M is a positive integer;
performing face recognition processing based on the obtained M levels of face sample sub-libraries,
wherein the dividing the password face sample library to obtain M levels of face sample sub-libraries according to the historical business information comprises:
dividing at least one password face sample library to obtain at least one level-1 face sample sub-library according to an attribute of a type of historical business information;
for each level-1 face sample sub-library, obtaining a number of face samples in the level-1 face sample sub-library, and under a condition that the number of face samples in the level-1 face sample sub-library is greater than or equal to the preset sample number threshold, dividing the level-1 face sample sub-library to obtain at least one level-2 face sample sub-library according to an attribute of another type of historical business information;
for each level-2 face sample sub-library, obtaining a number of face samples in the level-2 face sample sub-library, until a number of face samples in level-M face sample sub-library obtained by division according to an attribute of a certain type of historical business information is less than the preset sample number threshold.