Su Chang Chen

US Patent:

20220292654, Sep 15, 2022

Filed:

Mar 12, 2021

Appl. No.:

17/200338

Inventors:

- San Jose CA, US
Yifan Jiang - Austin TX, US
Yilin Wang - San Jose CA, US
Jianming Zhang - Campbell CA, US
Kalyan Sunkavalli - San Jose CA, US
Sarah Kong - Cupertino CA, US
Su Chen - San Jose CA, US
Sohrab Amirghodsi - Seattle WA, US
Zhe Lin - Fremont CA, US

International Classification:

G06T 5/50
G06T 7/194
G06T 11/60
G06T 11/00
G06N 3/04
G06N 3/08

Abstract:

The present disclosure relates to systems, methods, and non-transitory computer readable media for accurately, efficiently, and flexibly generating harmonized digital images utilizing a self-supervised image harmonization neural network. In particular, the disclosed systems can implement, and learn parameters for, a self-supervised image harmonization neural network to extract content from one digital image (disentangled from its appearance) and appearance from another from another digital image (disentangled from its content). For example, the disclosed systems can utilize a dual data augmentation method to generate diverse triplets for parameter learning (including input digital images, reference digital images, and pseudo ground truth digital images), via cropping a digital image with perturbations using three-dimensional color lookup tables (“LUTs”). Additionally, the disclosed systems can utilize the self-supervised image harmonization neural network to generate harmonized digital images that depict content from one digital image having the appearance of another digital image.