US 12,067,659 B2
Generating animated digital videos utilizing a character animation neural network informed by pose and motion embeddings
Yangtuanfeng Wang, London (GB); Duygu Ceylan Aksit, Mountain View, CA (US); Krishna Kumar Singh, San Jose, CA (US); and Niloy J Mitra, London (GB)
Assigned to Adobe Inc., San Jose, CA (US)
Filed by Adobe Inc., San Jose, CA (US)
Filed on Oct. 15, 2021, as Appl. No. 17/502,714.
Prior Publication US 2023/0123820 A1, Apr. 20, 2023
Int. Cl. G06T 13/40 (2011.01); G06N 3/045 (2023.01); G06N 3/08 (2023.01); G06N 3/088 (2023.01); G06T 7/20 (2017.01); G06T 7/73 (2017.01)
CPC G06T 13/40 (2013.01) [G06N 3/045 (2023.01); G06N 3/08 (2013.01); G06N 3/088 (2013.01); G06T 7/20 (2013.01); G06T 7/73 (2017.01); G06T 2207/10016 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/30196 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A non-transitory computer-readable medium comprising instructions that, when executed by at least one processor, cause a computing device to:
generate, utilizing a first neural network, a pose embedding from a pose in a sequence of digital poses;
generate, utilizing a second neural network, a motion embedding from the sequence of digital poses;
generate, utilizing a refinement neural network, a refined pose-motion embedding from the motion embedding and the pose embedding; and
generate, utilizing a generative neural network, a frame of a digital video depicting a character in motion utilizing inputs to the generative neural network comprising the refined pose-motion embedding and the motion embedding.