US 12,394,020 B2
Recommending objects for image composition using geometry-and-lighting aware search and efficient user interface workflows
Zhe Lin, Fremont, CA (US); Sijie Zhu, Orlando, FL (US); Jason Wen Yong Kuen, Santa Clara, CA (US); Scott Cohen, Sunnyvale, CA (US); and Zhifei Zhang, San Jose, CA (US)
Assigned to Adobe Inc., San Jose, CA (US)
Filed by Adobe Inc., San Jose, CA (US)
Filed on Apr. 11, 2022, as Appl. No. 17/658,774.
Prior Publication US 2023/0325992 A1, Oct. 12, 2023
Int. Cl. G06K 9/40 (2006.01); G06T 3/60 (2006.01); G06T 5/50 (2006.01); G06T 7/194 (2017.01)
CPC G06T 5/50 (2013.01) [G06T 3/60 (2013.01); G06T 7/194 (2017.01); G06T 2200/24 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/20221 (2013.01)] 20 Claims
OG exemplary drawing
 
8. A system comprising:
one or more memory devices comprising a geometry-lighting-aware neural network; and
one or more processors configured to cause the system to:
provide, for display within a graphical user interface of a client device, a background image for use in generating a composite image;
receive, via the graphical user interface of the client device, an indication to search for a foreground object image for the composite image; and
in response to receiving the indication to search for the foreground object image:
determine, utilizing the geometry-lighting-aware neural network, one or more foreground object images for the composite image;
generate a plurality of candidate locations for the one or more foreground object images within the background image by generating a grid of locations for the background image;
determine, from the plurality of candidate locations, a recommended location within the background image for a foreground object image from the one or more foreground object images based on determining similarity scores for the foreground object image and the locations for the background image from the grid;
generate the composite image utilizing the background image and the foreground object image from the one or more foreground object images at the recommended location; and
provide the composite image for display within the graphical user interface of the client device.