Definitive proof Lichtenstein is more creative than Da Vinci? This algorithm ranks great artworks by their creativity
The study scores 1,710 paintings, including works by Rodin, Michelangelo, Da Vinci and Mondrian, by their creativity (vertical axis) (Source: Arxiv.org)
It turns out Leonardo Da Vinci wasn't as much of a creative genius as we all thought – or that's according to an algorithm created by two computer scientists at Rutgers University in New Jersey, which reckons Mondrian and Liechtenstein's decidedly modernist works far outrank Da Vinci's various iterations of Madonna and Child.
The algorithm, created by Ahmed Elgammal and Babak Saleh, measures aspects of great paintings and sculptures, including space, texture, colour, harmony, proportion and pattern. "These artistic concepts can, more or less, be quantified by today's computer vision technology," they point out in their paper, Quantifying Creativity in Great Artworks.
Now, those of an artistic persuasion might suggest that quantifying creativity is a bit like trying to sieve water: whether something is creative is really a matter of opinion (cf. Michael Craig-Martin's 1973 work "An Oak Tree"). So what's the point of all this?
"The proposed framework would provide a ready-to-use approach that can utilise any future advances in computer vision that might provide better ways for visual quantification of digitised paintings," they suggest.
Munch's Der Schrei der Natur was one of the creative outliers picked out by the algorithm (Source: Getty)
"In fact, we applied the proposed framework using state-of-the-art computer vision techniques and achieved very reasonable automatic quantification of creativity on two large datasets of paintings," they add, hopefully. So there you go.
What was interesting was that although their algorithm didn't have any background on art history, it still managed to pick out some of the art world's most influential works, such as Edvard Munch's Der Schrei der Natur (aka The Scream), and Picasso's Les Demoiselle d'Avignon, as being the most creative.
Its limitations? Firstly, it can only analyse images of works posted online, meaning it's looking at a "closed set" of data. Secondly, it's looking at flat images – which means it can't assess, for instance, the exquisite detail in Michelangelo's David.
The authors also point out that it can't work out context.
"Clearly, it is not possible to judge creativity based on one specific aspect, eg. use of colour, perspective, subject matter, etc. For example, it was the use of perspective that characterised the creativity at certain points of art history, however it is not the same aspects for other periods." So they're not replacing subjectivity just yet.