Saturday, June 20, 2015

There must be a Pony in there somewhere


BRT has talked often about neural nets and their ability to learn by example, a notion that tends to be discussed in the abstract, until now.




In looking at these amazing images, one sees the beginnings of true AI, where interpreting reality via neural nets resembles, in indirect fashion, the paintings of Dali, Hieronymus Bosch or better yet, Chagall, where daydreams rule in unusual yet recognizable forms.


I and the Village - Chagall

When you let the system roam, images emerge similar to Van Gogh, electric, evocative and powerful.


The Starry Night - Van Gogh

If we apply the algorithm iteratively on its own outputs and apply some zooming after each iteration, we get an endless stream of new impressions, exploring the set of things the network knows about. We can even start this process from a random-noise image, so that the result becomes purely the result of the neural network, as seen in the following images:





No comments: