Are some people more / less susceptible to optical illusions?

[rowrrbazzle]

@ftg

I’ve generated RDS’s over the years myself. I notice some of my students had problems with seeing them so a took one and copied it onto a transparency and put it on a glass door. I’d tell the non-see-ers to look thru the door a bit. This helped a few people but not many.

That’s a great idea. But perhaps the following might make it more effective.

Person A who can see the image stands on one side of the door and views it until they see the 3-d.

Now position Person B on the other side of the door where Person A can see Person B normally while still seeing the 3-d image.

Then have Person A stand aside and have Person D (difficulty) stand exactly where Person A was.

Then tell Person D to look at Person B normally.

[Hermitian]

@rowrrbazzle

If you succeed, you’ll see that the above pattern now has the word “FAR” farther away than the word “NEAR”.

I can do magic eye illusions both ways (cross eyes or lengthening focus) and I cannot make your letters do anything.

[Colophon]

By coincidence, I was showing some optical illusions to my 7-year-old daughter just a couple of days ago, and she seems to be immune to most of them. All the “these lines aren’t parallel”, “these lines aren’t the same length” tricks totally failed to work on her - unless she was just being cussed to annoy me. A couple of the images that appear to show movement in a still image seemed to work on her, but not others.

Are children somehow less likely to be fooled by these things?
Edit to add: re Magic Eye images, how do people manage to do the eye crossing thing while still having the Magic Eye image in focus - likewise, looking through a glass door, wouldn’t the Magic Eye image be too blurry to work if you’re focusing on something beyond the glass? I can never get those things to work, as I have one very weak eye.

[Mama_Zappa]

@Mijin

They just explain it badly. You’re staring at a picture for ages waiting for some mystical thing to happen, when all that’s required is to consciously control the swivelling of your eyes. Here’s my best stab at it:

Can you cross your eyes (like the guy in this photo is doing)?
Do that now while looking at your screen. Notice that you can now see “two” screens (because the screen is in a different position in each eye’s vision and so your brain is not resolving them into one object).
Now, here’s the thing: you need to practice crossing your eyes a little. So the “two” screens are mostly overlapping.

Now take a look at a stereogram. Notice there’s always some kind of repeating tiling, like wallpaper. All you need to do is cross your eyes slightly, making the images overlap such that one “tile” overlaps its neighbour. At that point, the full 3D image pops out.

Crossing your eyes is an uncomfortable thing to do, but its just to learn the trick.
Once you’ve got it, crossing your eyes to view a regular stereogram will eventually feel just like the normal refocusing you do hundreds of times a day (because it is).

What finally did it for me was when a co-worker had one in a frame, with glass in front. I focused on my own reflection in the glass - and then had no trouble seeing the picture.

I’m out of the habit now; it would be hard to see them now.

[Colophon]

@Mama_Zappa

What finally did it for me was when a co-worker had one in a frame, with glass in front. I focused on my own reflection in the glass - and then had no trouble seeing the picture.

But if you focus on the reflection, isn’t the picture blurry? Or does it have to be blurry in order to work? In other words, can you ever actually see a Magic Eye picture “in sharp focus”, or will it always be a vague blurry blob?

[BigT]

@Colophon

But if you focus on the reflection, isn’t the picture blurry? Or does it have to be blurry in order to work? In other words, can you ever actually see a Magic Eye picture “in sharp focus”, or will it always be a vague blurry blob?

After I get the alignment right (both eyes looking at different parts of the image that correspond), I find that I automatically focus. That said, the illusion is all about the depth, not the individual details, so it doesn’t really matter if it is blurry.

[DCnDC]

@thelurkinghorror

That’s a spatial frequency thing and can be affected by visual acuity, but also how far away you’re standing from the screen. You can affect it by squinting.

I printed out and made one of those paper “hollow face dragons” and I can confirm that it doesn’t work when you look at it from a few feet away with both eyes, but if you either move 10+ feet away or close an eye (or look at it through your cell phone camera), presto! If I concentrate I can usually force myself to see it as it really is, then “relaxing” makes the face just pop out, it’s really freaky to see it “happen” in real-time.

[Bear_Nenno]

@Colophon

Magic Eye images, how do people manage to do the eye crossing thing while still having the Magic Eye image in focus - likewise, looking through a glass door, wouldn’t the Magic Eye image be too blurry to work if you’re focusing on something beyond the glass? I can never get those things to work, as I have one very weak eye.

That’s the point. Do NOT focus on the image. If you attempt to focus on the image, all you will see is the messy dots and blobs. You have to intentionally focus beyond the image. Once you force your eyes to adjust for a different focal distance, a new image (in 3D) will manifest itself. All this talk about “crossing your eyes” and such, is just a way to describe the process of changing the focus depth of your eyes. So if the image is about 18 inches from your face, make your eyes adjust as if you are trying to focus on an image that is several feet away. Then the illusion will take effect.

[DSeid]

@MichaelEmouse

Hollow-Face illusion - Wikipedia
A concave face shape is mistaken for a convex one (because humans are used to encountering people with convex faces). People with schizophrenia and, if I remember correctly, on psychedelics are less likely to falsely see it as convex. …

Yup.

Also as a somewhat discriminating feature in autistic spectrum disorders.

This method uses two layers of contrasting dots that move in opposite directions inside a narrow column.

Take a look at it below, and ask yourself – is the column rotating, or are they just two sheets of dots sliding across each other? …

… brains reflect the traits more associated with ASD should be more detail-focussed, a signature of the sensory processing wiring that emerges from autism. … those individuals who ranked higher on the survey [of ASD traits] tended to show pupil fluctuations that suggested they were focussing more on the illusion’s details than the picture of a rotating column.

Our brains are evolved to detect expected signals from noise and to literally create patterns, signals, the differing overlapping levels of top-down abstract concepts, out of what is always noisy bottom-up input. In nature that usually works to our information processing advantage.
FWIW no matter how hard I try and what “tricks” I use I cannot get those Magic Eye ones to work.

[DSeid]

@Colophon

By coincidence, I was showing some optical illusions to my 7-year-old daughter just a couple of days ago, and she seems to be immune to most of them. All the “these lines aren’t parallel”, “these lines aren’t the same length” tricks totally failed to work on her - unless she was just being cussed to annoy me. A couple of the images that appear to show movement in a still image seemed to work on her, but not others.

Are children somehow less likely to be fooled by these things?

Yes.

My guess is that it correlates with when they stop being so freakishly good at those memory matching games.

There are also cultural differences in how illusions are perceived.

[DSeid]

Sorry for the multipost but this is interesting stuff to me and I figured I would pass along some of what I am finding.

A related bit is AI’s performance in responding to what humans perceive as perceptual illusions.

The emphasis is from the perspective that in most real world circumstances “optical illusions are not a bug but a feature” …

That bit on how within layer feedback increases an AI’s ability to “perceive” the illusion.

This one on how difficult it is for the deep learning technique of the generative adversarial network, which has otherwise been so impressive, to succeed with illusions.

But this one showing that a deep neural network trained for prediction will identify the illusory motion as humans perceive it.