Seeing in 3-D and perceiving depth (called stereopsis) is a very complicated process. Each of your eyes sees a single 2-D image, but because our eyes are positioned to see that object from slightly different angles, our brains can use this disparity to perform some quick mental geometry and figure out where an object actually is in space. You can get an idea of this disparity between the images seen from each of your eyes with this quick test.
But our way is not the only way to see 3-D objects and perceive depth. In fact, over the years, we have found that more and more animals – cats, horses, birds, and even some insects - are capable of perceiving their world in three dimensions. Not all of them rely on the same mental algorithms we do. Learning more about which animals can see in 3-D and how their brains process this information can tell us more about why 3-D vision and depth perception has evolved in different groups of animals. It might even be able to give us new ways to teach machines to visualize 3-D objects.
Cephalopods (octopus, squids, cuttlefish, and nautiluses) have surprisingly similar eye anatomy to humans, despite the fact that we have drastically different evolutionary lineages and live in wildly different environments. But our brains and the way in which we process the information from those similar eyes, is far from the same. Scientists from the University of Cambridge and the University of Minnesota were curious to see how a cephalopod brain processes visual information, and how it compares to our own mental algorithms. So they outfitted a cuttlefish, a type of cephalopod, with 3-D glasses!
3-D glasses use polarized filters to show each of your eyes the same object from slightly different locations, tricking your brain into perceiving a 3-D object. With this approach, the scientists were able to test if cuttlefish altered their response to images of prey based on the disparity between images being shown to their right and left eyes. They discovered that cuttlefish do perceive 3-D images and use their depth perception to improve their hunting efficiency. However, the results from the study suggested that cuttlefish brains likely use a completely different algorithm to perceive depth.
The more we find out the difference in these visual algorithms could teach us a lot about how vision evolved across the animal kingdom and how we can make use of these algorithms for machine learning.