The next layers can detect more complex patterns such as grids and circles. As you move deeper into the network, the layers will detect complicated objects such as cars, houses, trees, and people.