Picasso, the greatest artist of the 20th century, and arguably the most influential ever, wanted to draw like a child. He said
“All children are artists. The problem is how to remain an artist once he grows up.”
This work shows it is possible for the computer to process a photograph into child-like art. The same system is also able to approximate the work of Joan Miró, cave-art and other forms of naive art. User (Art historians recognise Naive art as a particular genre, we have abuse their nomenclature, with apologies). The key to this is a data structure based on a graph of nodes and arcs.
In addition, we think it is interesting that the same data structure used for painting these Naive art works can be used to classify objects (into broad equivalence classes) independent of the depictive style. This is in contrast to most computer vision classifiers that classify photographs only.
X Bai, Y-Z. Song, A. Balikai and P.M. Hall, “Structure is a visual class invariant. Proceedings of the 2008 Joint IAPR International Workshop on Structural, Syntactic, and Statistical Pattern Recognition. LNCS vol 5342, 329 – 338.