US 12,271,996 B2
Image lighting transfer
Sudeep Siddheshwar Katakol, Ann Arbor, MI (US); Taesung Park, San Francisco, CA (US); Aliakbar Darabi, Newcastle, WA (US); Kevin Duarte, San Jose, CA (US); and Ryan Joe Murdock, American Fork, UT (US)
Assigned to ADOBE INC, San Jose, CA (US)
Filed by ADOBE INC., San Jose, CA (US)
Filed on Feb. 8, 2023, as Appl. No. 18/166,189.
Prior Publication US 2024/0265625 A1, Aug. 8, 2024
Int. Cl. G06T 15/50 (2011.01); G06T 11/60 (2006.01)
CPC G06T 15/506 (2013.01) [G06T 11/60 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method for training a generative adversarial network (GAN) to transfer lighting from a reference image to a source image, the method comprising:
receiving, by a user interface of an image editing apparatus, the source image and the reference image;
generating, by a lighting estimator of the image editing apparatus, a lighting vector from the reference image;
applying, by the image generating apparatus, features of the source image and the lighting vector to a generative network of the GAN to create a generated image;
applying, by the image generating apparatus, features of the reference image and the lighting vector to a discriminative network of the GAN to update weights of the discriminative network; and
applying, by the image generating apparatus, features of the generated image and the lighting vector to the discriminative network to update weights of the generative network.