How is it that we can almost instantly recognise any photo? The brain, it turns out, is highly sensitive to the overall structure of images, enabling it to distinguish in a fraction of a second whether a photo has been taken in a man-made environment. To do this, the brain employs a bit of simple arithmetic. This is the conclusion of a study conducted by cognitive neuroscientists at the University of Amsterdam (UvA), now published in the scientific journal Journal of Neuroscience.
The study involved having trial subjects look at hundreds of different photos and measuring their brain activity while they indicated whether the photos showed a natural or an artificial environment.
When presenting images of a street versus a forest, for example, the measurements showed considerable differences in brain activity within just a tenth of a second. These differences were found to be so reliable that it was even possible to predict trial subjects' answers based on that activity alone.
The researchers then developed a computer model to help pinpoint which specific information the brain uses to get a quick impression of the environment shown in a photo. Their simulation revealed that the brain uses a bit of simple arithmetic to judge the content of a scene: the statistical distribution of light and dark. Far less complicated than previous hypotheses, this model replaces them all.
Groen, I.I.A, Ghebreab, S., Prins, H., Lamme, V.A.F. & Scholte H.S., (2013). ‘From Image Statistics to Scene Gist: Evoked Neural Activity Reveals Transition from Low-Level Natural Image Structure to Scene Category’ in The Journal of Neuroscience 33(48), 18814-18824.