US 11,941,727 B2
Systems and methods for facial image generation
Saeid Motiian, San Francisco, CA (US); Wei-An Lin, San Jose, CA (US); and Shabnam Ghadar, Menlo Park, CA (US)
Assigned to ADOBE INC., San Jose, CA (US)
Filed by ADOBE INC., San Jose, CA (US)
Filed on Jul. 21, 2022, as Appl. No. 17/813,987.
Prior Publication US 2024/0037805 A1, Feb. 1, 2024
Int. Cl. G06T 11/00 (2006.01); G06V 40/16 (2022.01)
CPC G06T 11/00 (2013.01) [G06V 40/168 (2022.01); G06T 2200/24 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method for facial image generation, comprising:
receiving an image depicting a face, wherein the face has an identity non-related attribute and a first identity-related attribute;
encoding, using an attribute encoder, the image to obtain an identity non-related attribute vector in an identity non-related attribute vector space, wherein the identity non-related attribute vector represents the identity non-related attribute;
receiving an identity-related vector, wherein the identity-related vector represents a second identity-related attribute different from the first identity-related attribute, and wherein the second identity-related attribute is independent of the image;
generating, using a mapping network of a machine learning model, a modified latent vector in a latent vector space based on the identity non-related attribute vector and the identity-related vector; and
generating, using an image generation network, a modified image based on the modified latent vector, wherein the modified image depicts a face that has the identity non-related attribute and the second identity-related attribute.