Let’s set up an experiment. Open the camera app on your smartphone and start recording video. Place the screen directly in front of your eyes and try using these live footage to move around the room. Difficult and annoying, right? Even with modern technology, everything is too soapy in free movement. Colors blend and shapes are distorted.
At the same time, scientists say that this is very similar to how our eyes actually perceive the world around us. They are the same lens. Only worse, with biological defects. About the same chaotic visual data, our eyes constantly bombard our brains. Then, logically, the question arises: how do we see everything so clearly and without distortion?
This issue has been tried for many years. There are several theories that explain how our eyes and brain work together in an attempt to smooth out what we see around us. It is, first of all, a theoryblindness to change» (= when the stimulus changes, but we do not notice it, cutting off everything as superfluous to keep clarity). AND “blindness due to inattention» (= our inability to notice the visible object because our attention is focused on something else). These factors can somewhat explain our super-cool stabilization and lack of motion blur. And they have already proved very useful to us: these theories have inspired the creation of real technologies, such as algorithms for smoothing video on smartphones.
But in a study released a few weeks ago, scientists decided to test and prove a third theory, known as “serial addiction.” And it means that we are looking at an average of 15 seconds into the past.
In a new article published in January in the journal Science Advances, researchers from the University of Aberdeen and UC Berkeley describe a “previously unknown visual illusion.” It helps us understand how our brain smooths out the changes that occur over time.
The authors say:
Instead of analyzing each individual visual snapshot, our brains essentially form an arithmetic mean of what we have seen in the last 20 seconds. In this case, more weight is given to older frames.
Objects captured by the eye are brought closer together so that they appear more similar to each other. This is how our brain tricks us into perceiving our environment as more stable and still. Living “in the past” may explain why we don’t notice small changes that happen over time.
Here, let’s say, an illusion with objects in a room:
All objects are gradually changing, the room from the beginning and from the end of the video are very different. But since this happens gradually, in a gradient, we (on average) feel the room is unchanged. Arithmetic mean during these 25 seconds, it is the same.
For example, you might think about what happens to your eyeballs when they are focused on some object. When you move, even just while breathing, your eyes must move to keep the subject in focus. Therefore, objects that are “out of focus” quickly become blurred. We are used to it. But how does the main object stay sharp, do we move? Where do natural distortions go?
The researchers write in their paper:
Why do objects seem so stable to us, despite the constant changes in their projection on the retina?
Images on the retina constantly fluctuate due to many sources of internal and external noise – movement, occlusions, discontinuities, changes in lighting and perspective. But it seems to us that objects do not tremble, fluctuate, or change their identity from moment to moment. This question – why the world around us seems to change over time – has been around for centuries.
We hypothesize that this is due to an active mechanism in our brain whereby the representation of an object is constantly merging over time. A series of different images merge into one that, on average, best represents reality. And the consequence of this is the illusion of stability. In which the appearance of objects is shifted towards the past.
Our results provide a direct demonstration of the link between serial addiction and perceived visual stability in everyday life.
In general, “sequential dependency” causes objects to average out at any given moment and appear to be similar to those we have seen in the recent past. This creates a smoothing effect by reducing the total number of “frames” and reducing the total information that we need to absorb every second.
This explains visual illusions in a new way, which were not always possible to understand with the help of other theories.
Here, for example, is the famous video of a gorilla that goes unnoticed:
From the point of view of the new theory, the reason – in the arithmetic mean frame of the gorilla, as it were, does not exist. There are players and the ball, and the gorilla is too fleeting and insignificant to notice. We are not focused on her, and she is in the frame for less than eight seconds. If she were there for 20 seconds or more, then, according to the “sequential dependence”, our brain would always successfully recognize her, because she was present in the average frame.
The essence of the experiment
To test this theory, the researchers conducted an experiment. In it, different groups of people looked at a gradually changing image with an aging or younger face. According to scientists, if our brain really works on the principle of sequential dependence, it will fix the change in age with a lag. It is wrong to evaluate it, focusing on the recent past. And it will be possible to understand what kind of “time lag” is built into all of us.
First, two groups of 45 were shown a static image of a young face (13 years old) and an older face of the same girl (25.5 years old). Then they were asked to say how old she was in each photo. And in general they guessed right!
Then a third group of 47 people was shown a film with a face that changed, quickly becoming more mature. And they were asked to immediately estimate the age of the final face. It seemed to them much younger – the girl was given not 25.5 years, but 20.2 years.
Finally, for testing, the fourth group was given a face changing from old to young. And they asked them to estimate the age of a young face. It seemed to them much older than the real figure – 18.4 years instead of 13 years. Although, it would seem, in contrast, the girl should, on the contrary, seem even younger to them.
By the way, here is the video that was shown in the experiment:
In general, if the group was given just a photo of a person, they, on average, correctly guessed his age. But if the photo changed quickly, they calculated the arithmetic mean, moreover, with a time delay. And they gave the average age of the person who was seen 12-15 seconds ago.
To further test the relationship, the researchers increased the time intervals between image changes. The illusion of perception, an incorrect subjective assessment of age, persisted up to a delay of 15 seconds. A continuously changing physical object has been mistakenly perceived as unchanging by our previous visual experience.
Scientists say that this illusion of stability cannot be explained by passive “blindness to change.” And this is not “blindness due to inattention”: people really looked at this object, nothing else happened in the frame. Yet their feelings were distorted. They say that this is precisely “sequential dependence” – the perception of the following frames depends on what we saw earlier. And outwardly, objects seem to us the same as we saw them a few seconds ago.
In general, according to scientists from the University of Aberdeen and the University of California at Berkeley, we all constantly look at 12-15 seconds into the past, and due to this we get a high-quality picture and do not go crazy. It seems rather paradoxical – for example, I don’t feel it at all …