Human vision has evolved to recognize human figures in all different cases whether it is the real world or art through representations.
The situation becomes a bit more complicated when more abstract forms are involved art such as cubism or hyperrealism in which the shapes are altered, often making it difficult to recognize them. If people have proven that they do not have much difficulty in recognizing a person in a Picasso painting, for computers things do not seem to be so simple.
To this question, to what extent is a program computer μπορεί να αναγνωρίσει ανθρώπινες φιγούρες σε πίνακες ζωγραφικής που εντάσσονται στο κίνημα του κυβισμού προσπάθησαν να απαντήσουν ορισμένοι επιστήμονες του πανεπιστημίου του Μπέρκλι στην Καλιφόρνια των USA in their study "Recognizing People in Cubism."
Cubism was one of the most influential movements in 20th century painting, with pioneering painters such as Georges Braque and Pablo Picasso. In the images of the respective panels the figures are presented zigzagged, as if the optics that would have observers from different angles were joined. As a result, a cubic painting contains many fragments of the painter's perception of the same object.
The method followed by the researchers in their work was to select a list of 218 cubic tables with titles indicating that they contained human figures. They then asked 18 participants to rate the degree of abstraction in each image, on a scale from 1 to 5 and to form a rectangle around each human being.
Then they turned to their computers, trying to reproduce the results. For this purpose, they have used neural-based algorithms, a technique that in recent years has revolutionized facial recognition as it uses existing data to educate and improve its performance. Initially training their system with photos of people, they then fed it with the 218 database of tables they had presented to the survey participants, with the same requirement to place an outline around each person.
The results give a very useful look at how computational vision algorithms work and are also revealing to human perception. Without being surprised, people are better off in the cubic estimation, accumulating twice as much in relative assessment as the computers. As the removal in the table increases, the results for the algorithm become more and more destructive, unlike people who do not seem to be so affected. This observation contradicts a widespread perception of neurologists that people perceive objects by analyzing their individual parts
However, it seems that if computational vision can compete with human and even surpass it in normal conditions, the abstract conditions that art makes them show that human perception is much more flexible and reliable. However, the results of this research will be exploited to optimize computer vision, which one day not too far from today may be able to better assess a modern art table.
Source: naftemporiki.gr
