Art Imitations Spotted by Computers

by Wendelyn Bailey

According to the National Academy of Science, it is not just an amazing concept… this idea of a computer that is capable of recognizing bogus art… but a proven reality.

Established researchers have cleverly come up with a clear-cut technique that effectively creates the artist’s masterpiece then digitally breaks it down into small little sections, calling it "sparse coding." This system has ascertained that a true work of art has the capability of being reconstructed using these tiny sections whereas a copycat could not.

A little over a decade ago a comparatively innovative breakdown received attention universally through “drip paintings” of Jackson Pollock’s. However even after all this time it is stilled embroiled in much furor as several endeavors have been made using the method are still harvesting due to ambiguous findings.

Sparse coding brought about by Daniels Rockmore and numerous contemporaries are thought to be a lot more precise than those before.

How does this work? Digitally – and very easily. The first step to successfully testing a piece of work begins with slicing it into 144 segments then arranging them in rows and columns of twelve. Next, produce sets of one hundred forty-four indiscriminate elements the same size but in diverse assortment of forms. A computer will then modify these elements and arrange them to restructure the masterpiece. Again, keep in mind that an imitation will not be able to be reconstructed.

A computer, in the course of adjusting them will be capable of evolving different portions and areas of the original piece of art painting. It then goes one step further by making certain minuet; attainable amounts in that section are seized. This reconstruction is regarded as the "sparsest" set of reproductive operations.

This sparse coding was tried and tested on works of art that are often the attention of a lot of imitators. Continually proving, percentage wise, that the fakes were not able to be reconstructed. Sixteenth century painter Pieter Bruegel the Elder was one of these.

Eight genuine Bruegel pieces were virtually broken down and set along side fake pieces of work for patch comparisons. Seven out of eight cases formed accurate results in the ability of reconstructing the authentic art work and debunking the fake one.

Although sparse coding is highly acknowledged in many cases, it is somewhat restricted as there are still a lot of artistic styles that it can not process as well. One should mention that this type of process needs a fair sized collection of works from any one artist in order to adequately have the capability of catching imitations. Subject matter is also a factor and landscaping appears to be favored for this function.

Acknowledgment has been given that sparse coding is not always the answer in all situations and should not be considered as a replacement to other implements currently utilized. Regardless, adamancy is still maintained that in particular circumstances; it’s a valuable and effective addition.

And it is not just authentication that this formula is beneficial. According to Professor Rockmore, he feels it could possibly lend its analysis to other problem areas in the world of art as well.