CPC G06F 3/04845 (2013.01) [G06F 3/0482 (2013.01); G06T 7/11 (2017.01); G06V 20/40 (2022.01); G06V 40/161 (2022.01); H04N 5/265 (2013.01); H04N 5/2628 (2013.01); G06F 3/04817 (2013.01); G06T 2207/10016 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/20132 (2013.01); G06T 2207/30201 (2013.01)] | 17 Claims |
1. A method comprising:
accessing a frame;
detecting within the frame a first face having a first facial configuration and a second face having a second facial configuration;
segmenting the frame to generate a first face segment and a second face segment;
selecting, based on the first facial configuration, a first convolutional neural network trained to modify the first facial configuration to have a third facial configuration;
selecting, based on the second facial configuration, a second convolutional neural network trained to modify the second facial configuration to have the third facial configuration, wherein the first convolutional neural network is a first style transfer neural network trained to transfer images from the first facial configuration to the third facial configuration and the second convolutional neural network is a second style transfer neural network trained to transfer images from the second facial configuration to the third facial configuration; and
modifying, using the first convolutional neural network and the second convolutional neural network, the first face segment and the second face segment, respectively, to generate a first modified face segment and a second modified face segment.
|