The proliferation of deep learning and neural networks has done wonders for the tech world, but it's also created a more enjoyable and less functional trend: filtering videos through famous paintings. Neural networks analyze the visual qualities that make up the painting, and transfer this onto the film, as easy as adjusting an image filter on Photoshop. The latest, stunning, example comes from Bhautik Joshi, who used cubist art from Pablo Picasso to turn clips from 2001: A Space Odyssey into a vivid dream.
Joshi has past form on this: he's previously run parts of Blade Runner through a deep neural network tuned to Vincent van Gogh's Starry Night, and even filtered a number of GIFs of Donald Trump using images of food. The one below was...
via The Verge
