US 12,067,730 B2
Panoptic segmentation refinement network
Zhe Lin, Clyde Hill, WA (US); Simon Su Chen, San Jose, CA (US); Jason Wen-youg Kuen, Santa Clara, CA (US); and Bo Sun, Lyndhurst, NJ (US)
Assigned to Adobe Inc., San Jose, CA (US)
Filed by ADOBE INC., San Jose, CA (US)
Filed on Oct. 6, 2021, as Appl. No. 17/495,618.
Prior Publication US 2023/0104262 A1, Apr. 6, 2023
Int. Cl. G06K 9/00 (2022.01); G06N 3/08 (2023.01); G06T 7/11 (2017.01); G06T 7/194 (2017.01)
CPC G06T 7/194 (2017.01) [G06N 3/08 (2013.01); G06T 7/11 (2017.01)] 19 Claims
OG exemplary drawing
 
1. A computerized system, the system comprising:
one or more processors; and
computer storage memory having computer-executable instructions stored thereon which, when executed by the one or more processors, implement a method comprising:
receiving an input image;
deriving, via at least a first model, a first mask and a second mask, the first mask indicates a set of objects in the input image belonging to a first object class, the second mask defines each instance of the set of objects;
generating a feature map by concatenating one or more features from at least: the input image, the first mask, and the second mask;
based on the feature map, generating, via at least a second model, a third mask, the third mask indicates which pixels of the input image correspond to a foreground of the input image, the foreground excludes pixels corresponding to a background of the input image; and
based on the generating of the third mask, causing presentation of an output image associated with the input image, wherein the output image includes a fourth mask that defines a second instance of the set of objects, the second instance not being defined in the second mask.