These paintings remind me of Louis Wain's work when he was mentally ill.
Which makes me wonder, are these sophisticated neural nets mentally ill, and what would a course of therapy for them be like?
Inceptionism: Going Deeper into Neural Networks
21–30 of 166 posts
Re: Inceptionism: Going Deeper into Neural Networks
#22I'll be the first to say it. It looks like an acid/shroom trip.
Maybe there's something to do with how our brains interpret information differently when under the influence of psychoactive drugs.
I've been looking at Aldous Huxley's "Doors of Perception" and other psychonautic works recently and he hypothesizes that these sorts of drugs filter out the usual signals from the CNS that shut out the parts of perception that are not important for you to receive for survival.
It might be some great leap of armchair psychology, but I think we're due for another psychedelic revival, especially considering the new advances in synthetic psychedelics, legalization of more harmless recreational drugs, new tests in medical research using MDMA/LSD/Psilocybin, and the cultural shift away from the 'War on drugs'.
Re: Inceptionism: Going Deeper into Neural Networks
#23Re: Inceptionism: Going Deeper into Neural Networks
#24I would love to see what would come out of a network trained to recognize pornographic images using this technique. :)
I think it would look like a Picasso: https://en.wikipedia.org/wiki/Les_Demoiselles_d'Avignon
Re: Inceptionism: Going Deeper into Neural Networks
#25This is one of the most astounding things I've ever seen. Some of these images look positively like art. And not just art, but good art.
Re: Inceptionism: Going Deeper into Neural Networks
#26I would love to see what would come out of a network trained to recognize pornographic images using this technique. :)
I'm sure that would be extremely disturbing. It would probably look like H.R. Giger's art.
Re: Inceptionism: Going Deeper into Neural Networks
#27I'd really like to see what an Electric Sheep looks like. Maybe if they did a collaboration with the Android team?
Don't you mean the Tyrell corporation?
Re: Inceptionism: Going Deeper into Neural Networks
#28Am I the only person who is not entirely happy about the overuse of the pop-culture term 'inception' for everything that is remotely nested, recursive or strange-loop-like? In this paper, we will focus on an efficient deep neural network architecture for computer vision, codenamed Inception, which derives its name from the Network in network paper by Lin et al [12] in conjunction with the famous “we need to go deeper…
I haven't personally noticed any buzz-wordiness about that term lately. Maybe I'm just not looking at the same stuff as you.
Re: Inceptionism: Going Deeper into Neural Networks
#29The reason they look so 'fractal-like' (e.g. trippy!) is because they actually are fractals!
In the same way a normal fractal is a recursive application of some drawing function, this is a recursive application of different generation or "recognition -> generation" drawing functions built on top of the CNN.
So I believe that, given a random noise image, these networks don't generate the crazy trippy fractal patterns directly. Instead, that happens by feeding the generated image back to the network over and over again (with e.g. zooming in between).
Think of it a bit like a Rorschach test. But instead of ink blots, we'd use random noise and an artificial neural network. And instead of switching to the next Rorschach card after someone thinks they see a pattern, you continuously move the ink blot around until it looks more and more like the image the person thinks they see.
But because we're dealing with ink, and we're just randomly scattering it around, you'd start to see more and more of your original guess, or other recognized patterns, throughout the different parts of the scattered ink. Repeat this over and over again and you have these amazing fractals!
Re: Inceptionism: Going Deeper into Neural Networks
#30Very cool. I wonder if there's some example code on Github to generate images like this?