Artificial intelligence determined the age of a child by the movement of his eyes.

For quite a long time, scientists believed that the distribution of attention when viewing a visual series depends on

separation of the object from the background. It turned out that visual attention is primarily distributed to objects that are most important in meaning.

In an experiment led by Kirsten Dalrymplefrom the University of Minnesota, scientists decided to combine empirical data on the attention of children aged 1.5 to 2.5 years and, based on this, identify a number of factors that regulate attention in visual search.

The study involved 19 children inage 1.5 years and 22 children 2.5 years old. They were shown images of more than 3 seconds of various scenes, and the movements of their eyes were fixed with the help of a special eytreker, and scientists made heat maps of fixations.

The neural network was able to correctly determine the agechild in the images on which he looks, in 83% of cases. So scientists have shown that neural networks can analyze images and understand how human attention is distributed.