US 11,790,565 B2
Compressing image-to-image models with average smoothing
Jian Ren, Highland Park, NJ (US); Menglei Chai, Los Angeles, CA (US); Sergey Tulyakov, Santa Monica, CA (US); and Qing Jin, Palo Alto, CA (US)
Assigned to Snap Inc., Santa Monica, CA (US)
Filed by Jian Ren, Highland Park, NJ (US); Menglei Chai, Los Angeles, CA (US); Sergey Tulyakov, Santa Monica, CA (US); and Qing Jin, Palo Alto, CA (US)
Filed on Mar. 4, 2021, as Appl. No. 17/191,970.
Prior Publication US 2022/0292724 A1, Sep. 15, 2022
Int. Cl. G06T 9/00 (2006.01); G06N 3/08 (2023.01)
CPC G06T 9/002 (2013.01) [G06N 3/08 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method of operating a generative adversarial network (GAN), comprising:
receiving an image having learned parameters;
using an input class of the image to determine scaling and shifting parameters in a normalization layer; and
compressing the image using the determined scaling and shifting parameters by performing average smoothing between parameter layers and normalization layers to smooth abrupt boundaries where semantic information changes.