Entertainment

A Neuro-Network learned how to turn clumsy selfies into works of art

A Neuro-Network learned how to turn clumsy selfies into works of art

Artificial intelligence specializing in the generation of images will make a real difference. Fabian Stelzer, author of the first film to be drawn entirely by AI, has introduced a new tool called “Stable Diffusion. It is designed to redraw people’s selfies in a variety of styles – now no one will complain about looking ugly in a photo.

If you have ever wanted to see your face on a poster for a thrilling movie or on the cover of an exciting novel about adventures in unusual places, this neural network will help you realize it. The system doesn’t just “ennoble” photos of people, it redraws them in a given style, be it fantasy, space opera, noir, or something that can’t be described in ordinary words at all. As with AI, it interprets the task in its own way, so the result is always unpredictable.

The downside of Stable Diffusion is that it needs about a dozen different photos of one person to learn how to draw them accurately. And still the result will be doubtful, but the more photos the AI learns, the better the quality of its work will be. Surely this will please fans of selfies, especially those who are always unhappy with the way they look on them.