In work that could aid efforts to develop more brain-like computer vision systems, MIT neuroscientists have tricked the visual brain into confusing one object with another, thereby demonstrating that time teaches us how to recognize objects.

As you scan this visual scene (indicated with green circle), you spot a beaver out of the corner of your eye. As you glance towards it, the image is swapped for a monkey. Using analogous stimuli to produce swaps at specific locations in the visual field, MIT graduate student Nuo Li and professor James DiCarlo show that the brain starts to confuse different objects after a few hours exposure to this altered visual world. The confusion is exactly that which is expected if the brain uses temporal contiguity to teach it how to recognize objects. Read More...