Why are you better at recognizing upright faces? | Hacker News Reader

Why are you better at recognizing upright faces?

medicalxpress.com

26 pointswglb9 comments

A really good example of this effect can be seen in Adele's "upside down face" album cover – shown in both orientations (currently!) here: https://duckduckgo.com/?q=adele+upside-down+allbum&atb=v225-...

It's amazing how, despite being completely and utterly non-normal the right way up, it looks fine rotated by 180º and your brain just fills in the blanks quite happily.

recap: there was that pretty famous old study/experiment where people wore special glasses that flipped the world upside down visually. It didn't take very long for their brains to figure out how to "see" normally with this inverted information. Then things were able to switch back when they stopped wearing the glasses, with the footnote that after the experience they did experience some random scene flipping as if their brains had trouble sticking with an interpretation. I don't know if this study has ever been reproduced or attempted.

I bring it up to say, it would be interesting to test facial recognition in the same way that this piece is talking about with subjects wearing "those" glasses..

Hypothesis: if everything was upside down but faces were still upright, then your brain wouldn’t adapt to seeing the world upside down.

I always wanted to try that, but with the scene randomly flipping - no, thanks, no more

> I bring it up to say, it would be interesting to test facial recognition in the same way that this piece is talking about with subjects wearing "those" glasses..

Yep. They definitely shouldn't have twisted poor dude's neck 180 degrees just for the experiment.

This experiment has been reproduced in one way or another. Most significantly by Jan Degenaar, who spent many hours wearing "left to right inverting goggles", in his excellent and readable thesis "Perceptual engagement, a sensorimotor approach to phenomenal experience".

To my mind, the most important observation is that the scene did not "fully" invert "back to normal", but it was only visuals which Degenaar had appropriately re-learned which appeared "normal". For example, on the first day, walking down a street would make him see a car drive down the road to one side of him, but he'd hear it on the other side of him. Later on, the vision and the sound appeared on the same side, but if he looked at the number plate it was still "backwards". He had learned to locate sound while wearing inverting goggles, so the experience was "normal", but not learned to read while wearing inverting goggles, so that experience was still "strange".

Other papers, such as "The myth of upright vision" show that the change happens faster than can be attributed to "neural plasticity". https://scholar.google.com/scholar?cluster=19624886074317322...

> it is revealing that he does not do better with upright faces than inverted faces for face detection and face identity matching," Duchaine said.

> "The absence of an advantage for the face orientation that he's had more experience with suggests that our great sensitivity to upright faces results from both our greater experience with them and an evolved component that makes our visual system better tuned to upright faces than inverted faces."

This seems contradictory, or otherwise confusing from which perspective the orientation is being measured from.

It's the latter: An "upright face" is a face that has its chin below the forehead. For our eyes, those look like upright faces, but in Claudio's eyes they look like what inverted faces look to us.

So, because Claudio is not better at detecting upright faces than inverted faces, even though most of the faces in his life have been upright, that suggests an evolutionary component.

Im more interested in Claudio

Reply on news.ycombinator.com