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.

Date of Patent:
Nov. 22, 2022

Filed:

Nov. 25, 2020
Applicant:

Qnap Systems, Inc., New Taipei, TW;

Inventors:

Chun-Yen Chen, New Taipei, TW;

Chan-Cheng Liu, New Taipei, TW;

Ting-An Lin, New Taipei, TW;

Assignee:

QNAP SYSTEMS, INC., New Taipei, TW;

Attorney:
Primary Examiner:
Int. Cl.
CPC ...
G06V 40/16 (2022.01); G06K 9/62 (2022.01); G06N 3/04 (2006.01); G06N 3/08 (2006.01); G06V 40/50 (2022.01);
U.S. Cl.
CPC ...
G06V 40/172 (2022.01); G06K 9/6256 (2013.01); G06N 3/04 (2013.01); G06N 3/08 (2013.01); G06V 40/161 (2022.01); G06V 40/171 (2022.01); G06V 40/50 (2022.01);
Abstract

A method for collecting facial recognition data includes: locating a first face area from an Nth image frame; extracting a first facial feature defined with S factors; acquiring a second facial feature extracted from a second face area shown in an (N−1)th image frame at a corresponding position; determining whether the first face area is relevant to the second face area, and assigning to the first face area a tracing code; determining whether to store the first facial feature according to a similarity level of the first facial feature to existent data; storing and inputting the first facial feature into a neural network to generate an adjusted feature defined with T factors if the similarity level of the first facial feature to the existent data is not lower than a preset level, wherein T is not smaller than S; acquiring adjusted data generated by inputting the existent data into the neural network; determining whether the person is a registered one according to a similarity level of the adjusted feature to adjusted data; and using the stored first facial feature for training the neural network if the person is a registered one.


Find Patent Forward Citations

Loading…