What visual information does the brain cull?

We can't see everything...
13 August 2019


Person looking ahead



What sort of culling of visual information does the brain do?


Neuroscientist Duncan Astle tackled forum user Sazr's question...

Duncan - I think what he's getting at there is that the visual world outside of us, in front of us is incredibly rich and there's an information processing capacity problem. And there’s a couple of ways in which the information is simplified it as it moves through the system. So I’m just going to tell you about two of them. The first one is very early on - so in our retinas the layer at the back of the eyes has lots of photo sensitive receptors but they're very densely packed around a region of the centre called the fovea, and the fovea is only big enough to pick up your thumbnail at arm's length - if everyone sticks their arms out then you'll get a sense that it's actually very small. And so what your eye is doing is, it's moving around the environment very quickly in special type of eye movements called saccades, and it's updating information as it goes. Actually, much of what you perceive is actually slightly out of date from the last time that you made a saccade to that location. So there's lots of information the brain's just filling in the gaps really, from the last time around, but then when it arrives at the brain it's then culled again. And the reason we do that is because it's actually much more strategic to focus all of the resources on the information that's most relevant to the task at hand.

So there's a really nice experiment that someone did where they had participants watch a video of people playing football. And the task for the participants was to count how many times the ball touches someone's foot. Whilst they're watching this - and they're watching very closely and attending very closely to the ball and the foot - on the video someone comes on in a gorilla costume and runs around the pitch and runs up the camera and it turns out that actually very few participants are aware of ever having seen the gorilla. And that's because what their brain essentially is doing is biasing the allocation of resources so heavily to the task at hand - which is the balls and the feet - that it's screening out what it considers to be irrelevant information, in this case the gorilla. So there's lots of culling that goes on.

Chris - Peter?

Peter - Can I ask you how quickly it's rebuilding this image? If it's wandering around, are we talking milliseconds or tens of milliseconds... what happens if a child jumps into the road? Could you have missed it because at some point your brain was actually looking over there, your eyes were looking over there?

Duncan - Yeah. So for instance, one of the reasons why you shouldn't use your mobile phone while driving is because you're devoting your resources elsewhere. So there's what we call top down and bottom up attention. So if there's something that's relevant, like the task the participants were given, then you use top down systems in the brain to drive attention towards location. But you can also get bottom up attention - so if something very salient or surprising happens, then attention can immediately be captured to that location and there's a dynamic interplay between those two drivers.


Add a comment