Move-art-critics-Computer-algorithm-reveals-original-masterpiece and most complex art/artists

  • The algorithm analysed 63,000 paintings for their originality and influence
  • It analysed artist’s painting techniques, styles, colours and subject matter 
  • Edvard Munch’s The Scream and Pablo Picasso’s work scored highest
  • Creativity also seems to spike with abstract work at end of 20th Century 

It is a topic endlessly debated by the art world, but now a computer program has produced its own list of what it believes to be the most creative paintings ever produced.

The algorithm, developed by computer scientists at Rutgers University in New Jersey, analysed 63,000 paintings and sculptures produced through out history for their artistic features.

It ranked paintings such as The Scream by Edvard Munch and Pablo Picasso’s Ladies of Avignon as among the most creative ever produced, but abstract works by Fernando Caldau topped the results.

Masterpieces by artists including Monet, Michelangelo and Goya also scored highly while paintings by artists like Albrecht Dürer and Jean Auguste Dominique Ingres did less well.

The programme examined visual features including colour, texture, proportion, perspective and use of contrast as well as techniques like brush strokes and even the subject matter.

It then compared each painting to others produced during the same period to examine its originality.

While in some cases it agreed with art experts about the creativity of certain paintings, it also placed some highly valuable works low on its list.

THE HISTORY OF ART IS ORANGE

Orange is the most dominant colour used in artworks through out history.

Using computer analysis, researchers at the Karolinska Institute in Sweden analysed paintings available for download on the internet.

they then extracted the colour information from them and plotted it on a graph.

They found that orange is the most commonly used colour in art, but over time it has been becoming more blue.

The plot was based on 94,526 images for the years 1800-2000.

The graph shows a clear trend toward more blue paintings toward the end of the 20th Century, with all colours increasing except for orange.

Garden at les Lauves, by post-impressionist Paul Cezanne – who is considered to be one of the most important artists of the 20th century, was ranked particularly poorly.

Perhaps unsurprisingly the computer also singled out distinctive works that have come to typify more modern styles.

Abstract art by Kazimir Severinovich Malevich scored highly as did work by contemporary artists like Roy Lichetenstein.

Writing in a paper published on the open science website arxiv.org, Ahmed elgammal and Babak Saleh, who were the researchers behind the programme, said: ‘The most important conclusion of this work is that, when introduced with a large collection of paintings and sculptures, the algorithm can successfully highlight paintings that is considered creative – original and influential.

‘The algorithm achieved that without any knowledge about art or art history encoded in its input.

‘In most cases the results of the algorithm are pieces of art that art historians indeed highlight as innovative and influential.

Read more: http://www.dailymail.co.uk/sciencetech/article-3126302/Move-art-critics-Computer-algorithm-reveals-original-masterpieces-time.html#ixzz3dIePu7Pq
Follow us: @MailOnline on Twitter | DailyMail on Facebook

Norwegian expressionist Edvard Munch's scream, shown above, was ranked as one of the most creative paintings ever produced by the computer algorithm. It sold for $120 million when it was auctioned in 2012

Norwegian expressionist Edvard Munch’s scream, shown above, was ranked as one of the most creative paintings ever produced by the computer algorithm. It sold for $120 million when it was auctioned in 2012

This graph shows the creativity scores of paintings between 1850 and 1950. Artists like Van Gogh, Picasso and Malevich rank highly while some of Cezanne's work scored less well in terms of creativity

This graph shows the creativity scores of paintings between 1850 and 1950. Artists like Van Gogh, Picasso and Malevich rank highly while some of Cezanne’s work scored less well in terms of creativity

This analysis of 62,000 paintings from the Wikiart database shows works by Da Vinci scoring highly but more abstract pieces from the late 20th Century by Dorazio, Calhau and Landfield get the highest creativity scores

This analysis of 62,000 paintings from the Wikiart database shows works by Da Vinci scoring highly but more abstract pieces from the late 20th Century by Dorazio, Calhau and Landfield get the highest creativity scores

‘The algorithm achieved this assessment by visual analysis of paintings and considering their dates only.’

The researchers also found art shows peaks in creativity around the late 15th to early 16th century at the time of the High Renaissance period and in the late 19th and early 20th century.

There was also a significant increase in creativity in the second half of the 20th century.

They used the computer algorithm to examine art stored in the Wikiart online database.

They also applied the programme to specific types of paintings to see if it could distinguish particular genres and spot creativity.

When looking at religious paintings specifically, they found that artists like Francisco Goya and Salvador Dali produced particularly creative works of art.

The graph above shows the results of analysis of religious paintings with Marc Chagall scoring highest for creativity

Computer image analysis of paintings and drawings – The …

by DG Stork – ‎Cited by 13 – ‎Related articles

Computer image analysis of paintings and drawings: An introduction to the literature. David G. Stork. Ricoh Innovations, 2882 Sand Hill Road Suite 115, Menlo …

Gizmag

Oct 3, 2012 – Computer scientists have created an algorithm that enables a computer … has produced surprisingly accurate and expert results in art analysis.

Sep 27, 2012 – When fine art is concerned, or visual arts in general for that matter, complex cognitive functions are at play as the viewer analyze it. As you go …

Computer analysis of art – ACM Digital Library

dl.acm.org/citation.cfm?id…
Association for Computing Machinery
by L Shamir – ‎2012 – ‎Cited by 15 – ‎Related articles

Jul 1, 2012 – Analysis of visual art is a highly complex cognitive task that depends on the very many aspects of the art as well as complex brain connectivity, …

TechRadar

Jun 26, 2015 – Computers can now detect when an artist copies themselves … pair went on to use the technique to analyse a pair of near-identical paintings …

Advertisements