Classic Portraits that do NOT EXIST - artificial painting in 4 progressions [styleGAN3] скачать в хорошем качестве

Classic Portraits that do NOT EXIST - artificial painting in 4 progressions [styleGAN3]

00:00 - Creativity Factor 0.5* 04:56 - Creativity Factor 1* 12:01 - Creativity Factor 2* 19:39 - Creativity Factor 3* the higher the number, the more "dreaming" is allowed for the neural network. What you see are thousands of classical paintings, that do not exist. All characters are fully synthetic, drawn by a deep learning system. The machine has learned from 1336 classical-art paintings "what a classical portrait looks like". Now its able to create its own portraits. - infinite - The animation represents a seamless movement in the latent space of this neural network. neural network: .styleGAN3: https://github.com/NVlabs/stylegan3 dataset/model used: . MetDataset: https://github.com/NVlabs/metfaces-da... (Metropolitan Museum of Art Collection) operation, rendering: lucid layers studios thanks to patty smith for this great soundtrack with that archaic genesis-like vibe.