US 12,142,079 B2
Feature conversion learning device, authentication device, feature conversion learning method, authentication method, and recording medium
Masatsugu Ichino, Chofu (JP); Daisuke Uenoyama, Chofu (JP); Tsubasa Boura, Chofu (JP); Takahiro Toizumi, Tokyo (JP); Masato Tsukada, Tokyo (JP); and Yuka Ogino, Tokyo (JP)
Assigned to NEC CORPORATION, Tokyo (JP); and THE UNIVERSITY OF ELECTRO-COMMUNICATIONS, Tokyo (JP)
Appl. No. 17/636,074
Filed by NEC Corporation, Tokyo (JP); and The University of Electro-Communications, Chofu (JP)
PCT Filed Mar. 18, 2021, PCT No. PCT/JP2021/011192
§ 371(c)(1), (2) Date Feb. 17, 2022,
PCT Pub. No. WO2022/195819, PCT Pub. Date Sep. 22, 2022.
Prior Publication US 2023/0360440 A1, Nov. 9, 2023
Int. Cl. G06V 40/18 (2022.01); G06T 3/40 (2006.01); G06V 10/77 (2022.01); G06V 10/774 (2022.01); G06V 10/776 (2022.01)
CPC G06V 40/193 (2022.01) [G06T 3/40 (2013.01); G06V 10/7715 (2022.01); G06V 10/774 (2022.01); G06V 10/776 (2022.01)] 9 Claims
OG exemplary drawing
 
1. An authentication device comprising:
a memory configured to store instructions; and
a processor configured to execute the instructions to:
acquire an authentication target image;
enlarge the authentication target image;
extract a feature of an image obtained by enlarging the authentication target image in a method that is the same as a feature extraction method for a comparative image having higher resolution than the authentication target image;
convert the feature of the image obtained by enlarging the authentication target image based on a result of learning a feature conversion method using a loss function whose value decreases as a difference between a feature after conversion from a feature of a learning image and a feature of a deteriorated image enlarged after the learning image is reduced decreases; and
compare a feature after conversion with a feature of the comparative image.