US 12,148,202 B2
Domain changes in generative adversarial networks
Sergey Demyanov, Santa Monica, CA (US); Konstantin Gudkov, London (GB); Fedor Zhdanov, London (GB); and Andrei Zharkov, London (GB)
Assigned to Snap Inc., Santa Monica, CA (US)
Filed by Snap Inc., Santa Monica, CA (US)
Filed on Jun. 15, 2022, as Appl. No. 17/841,333.
Prior Publication US 2023/0410479 A1, Dec. 21, 2023
Int. Cl. G06T 11/00 (2006.01); G06T 3/18 (2024.01); G06T 7/68 (2017.01); G06V 10/74 (2022.01); G06V 10/774 (2022.01); G06V 10/82 (2022.01); G06V 40/16 (2022.01)
CPC G06V 10/774 (2022.01) [G06T 3/18 (2024.01); G06T 7/68 (2017.01); G06T 11/00 (2013.01); G06V 10/74 (2022.01); G06V 10/82 (2022.01); G06V 40/16 (2022.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/20221 (2013.01); G06T 2207/30201 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
generating, using a pre-trained generative adversarial network (GAN), a plurality of images, the pre-trained GAN trained on a first plurality of target images;
determining a feature for each of the plurality of images;
determining the feature for each of a second plurality of target images;
matching, based on the feature, target images of the second plurality of target images with the plurality of images to generate a plurality of matched target images; and
training a discriminator of the pre-trained GAN with the plurality of matched target images and the plurality of images.