JeremyR
I was trawling the internet searching for images on machine learning and came across an artist by the name of Mondrian.
I heard of this guy back in my days of going to art college and you will probably recognise his work of his geometric shapes and use of primary colors and lines - a bit tedious to look at really.

But in the attachment you can see his artistic progression over some years with his work starting from impressionism through to cubism and into what is popular today - his abstract art. What is fascinating for me today is how he was able to extract out the essence of what he saw and put that in paint rather than spend hours on end putting in life's detail.

Im not sure how this relates to machine learning but perhaps its that AI is to recognise whats familiar without getting too caught up in the detail and create a so-called intuitive response based on experience, just like Mondrian did over 100 years ago! If you look at his final abstract style it looks a lot like data representation that software pumps out today.

Anyway food for thought :) Click image for larger version - Name: mondrian_evolution.jpg, Views: 25, Size: 836.76 KB
Quote 0 0
pshipkov
There is research and software libraries out there that deal already with stuff like this.
The detail regression of the artist is pretty extreme.

I am sure something like this can be achieved without the need of neural nets.
Quote 0 0
JeremyR
Yeah Its just a post on observation not necessarily a wish list idea or anything. Its a comment on presenting and analysing data, since machine learning and AI is getting really big these days for companies not related to animation.

Big companies like Google, Apple and Amazon are investing heavily in machine learning, we have all this data about people and only now being able to throw it all in one bucket for computers to look at and make its own diagnosis on things. 
Quote 0 0

Add a Website Forum to your website.