Interesting article, and makes a good point that there could be different ways to get good at something.
Just curious, has anyone tried to train people to be good at face recognition in the way "super-recognizers" are good? I imagine there are practical advantages to forensic face recognition (e.g. training, repeatability, etc.), so maybe no one has tried. But I wonder if super-recognition is pure talent, or if there could be two totally separate ways of getting good at face recognition. Like for acting, method vs classical, or for music, Suzuki method for music vs sheet music, etc.
Just curious, has anyone tried to train people to be good at face recognition in the way "super-recognizers" are good? I imagine there are practical advantages to forensic face recognition (e.g. training, repeatability, etc.), so maybe no one has tried. But I wonder if super-recognition is pure talent, or if there could be two totally separate ways of getting good at face recognition. Like for acting, method vs classical, or for music, Suzuki method for music vs sheet music, etc.
My intuition is telling me that facial recognition is different from what chess masters do. It seems that someone who is good at facial recognition would be able to ignore "noise", (which was part of the Cambridge test), like with a beard or without a beard. Whereas chess masters very often point out things like "position x is just like position y except that the pawn is on a rook file..." Chess recognition seems much more dynamic than facial recognition. Maybe I'm oversimplifying facial recognition.
I got an 86% accuracy on the Cambridge test, which is better than my chess accuracy.
I got an 86% accuracy on the Cambridge test, which is better than my chess accuracy.
<Comment deleted by user>
<Comment deleted by user>
> .... I want to think through with you here.
me: brilliant move!
The problem with only being concerned with experts experimentally and in the theory of chess thinking, is how it might be used to infer about theory of learning for chess. Besides, the lack of "dynamic range" in the expermental design. If one wants to get more information about a multidimensional unknown function, probing it on a small cluster intensively, might not tell us what makes it tick... Not sure it is, even for the target question of mere thinking theory or model, an efficient search strategy. Maybe, I have been exposed to black box puzzle and perturbation signals as a way to look at response diversity as proving inducible information from the insides. But I think is a general analytical instrument of measure principle, to have calibrations scales. Maybe I assume a sort of continuum of experiences where there is information to be gotten from having more regularly spread data points..
But yes, if one is clear about the target question being how experts think, well start by focusing on them.. of course.
It may not be the only interesting scientific question about chess and how human thinking about it, or more interestingly of one can learn from individual play experience, or other means of getting to know about how the board behaves, from others recounting in various shared resources. or even good old human story telling to each other about many heads past discoveries. etc...
me: brilliant move!
The problem with only being concerned with experts experimentally and in the theory of chess thinking, is how it might be used to infer about theory of learning for chess. Besides, the lack of "dynamic range" in the expermental design. If one wants to get more information about a multidimensional unknown function, probing it on a small cluster intensively, might not tell us what makes it tick... Not sure it is, even for the target question of mere thinking theory or model, an efficient search strategy. Maybe, I have been exposed to black box puzzle and perturbation signals as a way to look at response diversity as proving inducible information from the insides. But I think is a general analytical instrument of measure principle, to have calibrations scales. Maybe I assume a sort of continuum of experiences where there is information to be gotten from having more regularly spread data points..
But yes, if one is clear about the target question being how experts think, well start by focusing on them.. of course.
It may not be the only interesting scientific question about chess and how human thinking about it, or more interestingly of one can learn from individual play experience, or other means of getting to know about how the board behaves, from others recounting in various shared resources. or even good old human story telling to each other about many heads past discoveries. etc...
I there a known interaction between our own body map, and how we project intention from watching the others posture or facial expression or the eye gaze vector?
I have pops up that keep connecting things I read with things I think I know or knew. And kind of having read that even between mammals there is some degree cross body mapping attribution (I think even dolphins might new they human gesticulating side-kicks as dolphins mapping arms and lower body to their own. Like they are down reverse anthropomorphism.
I did not need that, but it is a theme. Also other thoughts lurking in the direction of chess and our body internal models or maps.
So I was thinking, about such specific areas for face recognition not being about general object recognition 3D space transformation robustness of recognition. Is there any relation to looking at others faces and perhaps our body map. I am not good about atlas and names.. perhaps with frequency I might know occipital from temporal, but then parietal, got me, it is not frontal or the others. what might it be.. (not important). The important is that somewhere the is a known brain anatomical coordinate association to a function of high level observable. (lack English words flexibility at times).
Or maybe you have some other thoughts that might make this kind of association/question of mine, not even smelling likely relatable. Body map, intent projection abilities, and other empathy types, and facial recognition.
Another pop pup question while reading:
One thing that stands out, which I find might have something to think about in relation to chess, is the nature of the feature importance to our social needs (human social animal, subconscious I bet).
I find that in chess theory some definitions of concepts of feature on the board to be concerned with, require for being applicable that the feature be also carrying an advantage value, that if it is not so, it would not be that named feature. I may be about using categorical levels for features, there or not, and then if not having degrees to share, well, an approximation would be to make a threshold under which it is brain load wasted space (and we know there is limited resource there, in performance mode and time pressure).
It might be not just chess theory, but how the author coining or choosing that the feature might not be worth an attentional ressource chunk (units? :)) in own authors internal pattern recognition through experience. If I don't make any sense, that is all right. I have lots of those questions.... my way of sustaining my labile attentional resources. I use my curiosity trait as carrot.
Tying back to why I went there though, is that a person's face, on a picture is likely, in our social animal evolution, to be a real human being in our face as well, and I would think our "biological" view of the world would make it more important than a less distinguishable class of objects the purpose or importance to own "happiness" (some might jump to "survival" a bit dramatic), to differentiate a chair from another, as long as one can sit on it.. putting upside down, does not need us to also known if another upside down also have different types of nails or shape. Not sure I made a point here.. maybe you can find one.
I have pops up that keep connecting things I read with things I think I know or knew. And kind of having read that even between mammals there is some degree cross body mapping attribution (I think even dolphins might new they human gesticulating side-kicks as dolphins mapping arms and lower body to their own. Like they are down reverse anthropomorphism.
I did not need that, but it is a theme. Also other thoughts lurking in the direction of chess and our body internal models or maps.
So I was thinking, about such specific areas for face recognition not being about general object recognition 3D space transformation robustness of recognition. Is there any relation to looking at others faces and perhaps our body map. I am not good about atlas and names.. perhaps with frequency I might know occipital from temporal, but then parietal, got me, it is not frontal or the others. what might it be.. (not important). The important is that somewhere the is a known brain anatomical coordinate association to a function of high level observable. (lack English words flexibility at times).
Or maybe you have some other thoughts that might make this kind of association/question of mine, not even smelling likely relatable. Body map, intent projection abilities, and other empathy types, and facial recognition.
Another pop pup question while reading:
One thing that stands out, which I find might have something to think about in relation to chess, is the nature of the feature importance to our social needs (human social animal, subconscious I bet).
I find that in chess theory some definitions of concepts of feature on the board to be concerned with, require for being applicable that the feature be also carrying an advantage value, that if it is not so, it would not be that named feature. I may be about using categorical levels for features, there or not, and then if not having degrees to share, well, an approximation would be to make a threshold under which it is brain load wasted space (and we know there is limited resource there, in performance mode and time pressure).
It might be not just chess theory, but how the author coining or choosing that the feature might not be worth an attentional ressource chunk (units? :)) in own authors internal pattern recognition through experience. If I don't make any sense, that is all right. I have lots of those questions.... my way of sustaining my labile attentional resources. I use my curiosity trait as carrot.
Tying back to why I went there though, is that a person's face, on a picture is likely, in our social animal evolution, to be a real human being in our face as well, and I would think our "biological" view of the world would make it more important than a less distinguishable class of objects the purpose or importance to own "happiness" (some might jump to "survival" a bit dramatic), to differentiate a chair from another, as long as one can sit on it.. putting upside down, does not need us to also known if another upside down also have different types of nails or shape. Not sure I made a point here.. maybe you can find one.
There is also an assumptions maybe that all experts are maximal experts over the same coverage of chess positions.
It might actually be likely so, if contemporary, and having had good mixing pairing among many of each others, augmenting the chances of serving each other complementary experience sets or familiarities.
But even if they were having expertise, in the end, that would averagely cover the union of all the chess world coverage, the initial conditions and individual sub-optimal expertise while they were not yet experts over the widest domain, are likely to have had pattern possible emerge in one their internal radar with different "landscape" boundaries.. I might be skipping a lot of arguments here, but I hope these words finds you well enough... if not we can think together some more here.. The words of chess theory, might serve a role of adjusting toward a common convention of pattern discovery in ones learning.
I guess the impact of the only one dimensional measure of chess progress in the rating might also have an influence in how people look at chess thinking and learning. In spite of possibly diverse experiences set trajectories, among all the current experts. And past what chess theory could guide, high level chess does a lot of in between category hunch optimization, and probably a lot of subtle subjective patterns formation in internal representations of chess to recognize in other future experience, none of which they would have the time, energy or motive to share the theory of in words, unless they somehow enter a new phase of their lives, when the urgencies seems to be old pursuits, and they wonder about what will remain besides the record of their games... :)
so a big chess world, and human individual random walks might harly make for all experts to have internal patterns that are clones of each other. and that they might not have encountered those in the same order any way..by order I mean individual order, not that the chess world itself would necessarily have a "linear" order.
It might actually be likely so, if contemporary, and having had good mixing pairing among many of each others, augmenting the chances of serving each other complementary experience sets or familiarities.
But even if they were having expertise, in the end, that would averagely cover the union of all the chess world coverage, the initial conditions and individual sub-optimal expertise while they were not yet experts over the widest domain, are likely to have had pattern possible emerge in one their internal radar with different "landscape" boundaries.. I might be skipping a lot of arguments here, but I hope these words finds you well enough... if not we can think together some more here.. The words of chess theory, might serve a role of adjusting toward a common convention of pattern discovery in ones learning.
I guess the impact of the only one dimensional measure of chess progress in the rating might also have an influence in how people look at chess thinking and learning. In spite of possibly diverse experiences set trajectories, among all the current experts. And past what chess theory could guide, high level chess does a lot of in between category hunch optimization, and probably a lot of subtle subjective patterns formation in internal representations of chess to recognize in other future experience, none of which they would have the time, energy or motive to share the theory of in words, unless they somehow enter a new phase of their lives, when the urgencies seems to be old pursuits, and they wonder about what will remain besides the record of their games... :)
so a big chess world, and human individual random walks might harly make for all experts to have internal patterns that are clones of each other. and that they might not have encountered those in the same order any way..by order I mean individual order, not that the chess world itself would necessarily have a "linear" order.
what is phi(z), psi(z) and T(z) intended to illustrate or represent. I only undertand the symetric looking distribution as a function of z. and other functions of z as areas under the curve as function of the interval lower or upper variable bound, also label z. I feel untethered a bit between those those figures.. Is that a vision trick (kdding!).
@Graque said in #2:
> Interesting article, and makes a good point that there could be different ways to get good at something.
>
> Just curious, has anyone tried to train people to be good at face recognition in the way "super-recognizers" are good? I imagine there are practical advantages to forensic face recognition (e.g. training, repeatability, etc.), so maybe no one has tried. But I wonder if super-recognition is pure talent, or if there could be two totally separate ways of getting good at face recognition. Like for acting, method vs classical, or for music, Suzuki method for music vs sheet music, etc.
Sorry to be slow to respond here! There are definitely folks working on how to develop training regimens to improve face recognition ability, but my understanding is that it's quite hard to get much of a boost. In some ways, this is sensible given the big difference between familiar and unfamiliar face discrimination and matching - even though most people are very good at recognizing familiar faces, unfamiliar faces still lead to many matching errors. That sort of suggests that the expertise you have for faces you've seen a lot doesn't even transfer well to other faces you have yet to learn.
Some good names to follow up with about training and faces are Joe Degutis and Meike Ramon if you're interested in this. There is also a decent literature about trying to train up face recognition abilities in individuals with face blindness.
> Interesting article, and makes a good point that there could be different ways to get good at something.
>
> Just curious, has anyone tried to train people to be good at face recognition in the way "super-recognizers" are good? I imagine there are practical advantages to forensic face recognition (e.g. training, repeatability, etc.), so maybe no one has tried. But I wonder if super-recognition is pure talent, or if there could be two totally separate ways of getting good at face recognition. Like for acting, method vs classical, or for music, Suzuki method for music vs sheet music, etc.
Sorry to be slow to respond here! There are definitely folks working on how to develop training regimens to improve face recognition ability, but my understanding is that it's quite hard to get much of a boost. In some ways, this is sensible given the big difference between familiar and unfamiliar face discrimination and matching - even though most people are very good at recognizing familiar faces, unfamiliar faces still lead to many matching errors. That sort of suggests that the expertise you have for faces you've seen a lot doesn't even transfer well to other faces you have yet to learn.
Some good names to follow up with about training and faces are Joe Degutis and Meike Ramon if you're interested in this. There is also a decent literature about trying to train up face recognition abilities in individuals with face blindness.