Change Detection Demo

Animals not only capture attention, but they are monitored in an ongoing manner by a high-level, category-specialized system that was shaped by ancestral selection pressures, not general learning processes.

Demonstration of change detection:
Look to see if there are any changes in the scene

Demo 1

Demo 2

Even though the elephant blends into the savannah background and is smaller (in pixels) than the bright red minivan, changes to the elephant were detected 100% of the time, but changes to the red minivan were detected only 72% of the time.  This is a very typical result: Changes to nonhuman animals were detected faster and more accurately than changes to plants, tools, buildings, and vehicles.  That change detection is better for animals is particularly interesting.  If category-specialized attentional systems developed only as a result of domain-general learning and expertise, then one would predict better change detection for vehicles than for nonhuman animals.  As pedestrians and drivers, our lives depend on detecting changes in the state and location of vehicles, and we are trained on this on a daily basis.  Yet people are better at detecting changes to nonhuman animals.  A visual attention system specialized for monitoring animals was useful for our hunter-gatherer ancestors, but more of a distraction today.

Now try these:

Demo 3

Demo 4

Demo 5

Demo 6