US 11,055,514 B1
Image face manipulation
Chen Cao, Los Angeles, CA (US); Sergey Tulyakov, Marina del Rey, CA (US); and Zhenglin Geng, Santa Monica, CA (US)
Assigned to Snap Inc., Santa Monica, CA (US)
Filed by Snap Inc., Santa Monica, CA (US)
Filed on Dec. 14, 2018, as Appl. No. 16/220,859.
Int. Cl. G06K 9/00 (2006.01)
CPC G06K 9/00288 (2013.01) [G06K 9/00228 (2013.01); G06K 9/00268 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
receiving, by one or more processors, an input image comprising a face having a first expression;
obtaining, by the one or more processors, a target expression for the face, extracting, by the one or more processors, a texture of the face and a shape of the face;
generating, by the one or more processors based on the extracted texture of the face, a target texture corresponding to the obtained target expression using a first machine learning technique;
generating, by the one or more processors based on the extracted shape of the face, a target shape corresponding to the obtained target expression using a second machine learning technique, wherein coefficients of the first and second machine learning techniques are fixed while a third machine learning technique is trained; and
combining, by the one or more processors based on an output of the third machine learning technique, the generated target texture output of the first machine learning technique and generated target shape output of the second machine learning technique into an output image comprising the face having a second expression corresponding to the obtained target expression, wherein the output image comprising the face having the second expression corresponding to the obtained target expression is generated for display in a user interface.