This experiment uses three open-source machine learning models to generate an image.
First, a model is used to recognize faces in an image.
Then, two modern neural networks use that input as the base to synthesize a new image.
Pix2PixHD turns the recognize faces into new faces. It does this after being trained on a dataset of portraits.
Finally, a model called StyleGAN , trained on landscape images, transforms your background.

• 1Camera Image
• 2Recognize Faces
• 3Synthetic Portrait