There's a fascinating animated movie "Loving Vicent" coming up that is made in the style of Vincent van Gogh. Apparently the individual frames are actual oil paintings made by hand.

http://www.thisiscolossal.com/2016/02/loving-vincent-trailer...

When I first saw the trailer my first thought was the neural network approach shown here. I wonder how close you could get to "Loving Vincent" style by simply applying this algorithm to individual video camera frames.