Are You Brave Enough to Post A Picture of Yourself?

[ImAllLikeOkWaitWat]

this is me...

You are very beautiful.

 

[timewerx]

I do not think I agree with this assessment. As I said before, the program seems to be designed to assess white people. So you should not think of the program as biased towards white people in the sense that it believes white people are most attractive. It is biased in that the criteria it uses for judgment is only useful in judging/describing ideal white people.

To illustrate:
To say that I have a bonus in scores for being a white male would be like entering Miss America and telling the women they have a huge bonus in scores for being female.
Or like an orange telling an apple that it has a huge bonus in scores for being an apple in a program that assesses the quality of apples. (or perhaps a Granny Smith apple telling a Fuji apple...)
the program is not rating me more highly because I am a white male. It is supposed to be rating me more highly because the proportions of my face are close to the ideal of a white male.
People of other races are judged by a different set of proportions. just like a Granny Smith apple is judged by a different set of qualities than a Fuji apple.

I would argue that the results of this program, even if they are meaningful to white people, would be as meaningless in its application to you as my judging a Fuji apple as less attractive because it is not green like my Granny Smith apple criteria claims a Granny Smith apple should be.

An asian woman who scores highly in this program probably looks very much like a white person (proportion-wise). If you score lowly in this assessment. it does not mean you are an unattractive person, it just means you are an unattractive white person. But you might be a very attractive asian person.

I get what you mean that the program might be judging me as a white person against the white standard of beauty which maybe inappropriate standard for my case.

The one I used before, Howhot.io, it's gone already, they had diverse races in their database, including Asians. I even scored consistently lower in that one. I think Haystack.ai also had diverse racial profiles based on the sample pictures on their website.

The usual method of building a database for this kind of program is have random people given sets of pictures and rate them 1-10 in terms of attractiveness and the raw data inputted to the program's database.

This would only mean that most people (regardless of race) would find white people the most attractive....Based on my own observation, I think there's a lot of truth in it.

 

[ImAllLikeOkWaitWat]

This would only mean that most people (regardless of race) would find white people the most attractive....Based on my own observation, I think there's a lot of truth in it.

So there is a most attractive race after all eh?

 

[timewerx]

So there is a most attractive race after all eh?

"AI" programs like these still depend on surveys and surveys still come from people.

Results are still heavily influenced by human factor or what most people think is most beautiful.

The only benefit of these programs is you can see how you compare against the status quo....

However, I can assure you it is not universally true as perception of beauty isn't the same for everyone

 

[Grandpa2390]

"AI" programs like these still depend on surveys and surveys still come from people.

Results are still heavily influenced by human factor or what most people think is most beautiful.

The only benefit of these programs is you can see how you compare against the status quo....

However, I can assure you it is not universally true as perception of beauty isn't the same for everyone

The usual method of building a database for this kind of program is have random people given sets of pictures and rate them 1-10 in terms of attractiveness and the raw data inputted to the program's database.

This would only mean that most people (regardless of race) would find white people the most attractive....Based on my own observation, I think there's a lot of truth in it.

I don't think so. Scientists have been able to figure out that beauty is the average of features. I came across this website where a person determined the ideal female by each country in the world (choosing a majority race of course) by averaging a large sample of faces in that country's citizens. and the results are... exactly what we would expect. averaging 1000 caucasian faces gives you a 10/10 beautiful woman. and then you can measure the proportions (eyes, jaw, etc) and use that in a program to judge white faces. to see how closely they match the statistical average.

check this out:
https://jezebel.com/5758201/see-the-average-face-of-women-from-40-different-countries/

this mentions flaws such as racially biased and age biased. but I think those make sense. I don't think a composite of white faces should be chosen for South Africa, etc. you get what I mean though. A composite of a individual races.
As for the symmetry and lack of blemishes, they explain this as being due to the blending process and if you think about it, a more attractive face will lack blemishes and be more symmetrical.

I don't provide this as concrete evidence of what each nations view of attractive people is. I provide it as evidence that if you take a composite of many faces from a race, you will get a very attractive image. and you can see the proportions vary from a composite of white people faces to a composite of asian people faces. or black people faces, etc.

 

[Grandpa2390]

oops...wrong pic..lols..here's the right one...

So you are the first one or this one?

Both pictures are of very attractive ladies btw
though the movement in the first one freaked me out a bit at first lol.

 

[NurseAbigail]

So you are the first one or this one?

Both pictures are of very attractive ladies btw
though the movement in the first one freaked me out a bit at first lol.

lol i luv her expression, i'm the second one

 

[Ada Lovelace]

Fooling around with the "world's best facial recognition app of 2017" with a touched up picture of mine that my niece did (she added make up, wig, etc).

When I couldn't sleep last night I checked out Haystack. Apparently the "world's best facial recognition app of 2017" thinks I'm a happy 11-year-old. Don't know whether to laugh or cry about that, so I guess I'll go with .

 

[timewerx]

I don't think so. Scientists have been able to figure out that beauty is the average of features. I came across this website where a person determined the ideal female by each country in the world (choosing a majority race of course) by averaging a large sample of faces in that country's citizens. and the results are... exactly what we would expect. averaging 1000 caucasian faces gives you a 10/10 beautiful woman. and then you can measure the proportions (eyes, jaw, etc) and use that in a program to judge white faces. to see how closely they match the statistical average.

check this out:
https://jezebel.com/5758201/see-the-average-face-of-women-from-40-different-countries/

this mentions flaws such as racially biased and age biased. but I think those make sense. I don't think a composite of white faces should be chosen for South Africa, etc. you get what I mean though. A composite of a individual races.
As for the symmetry and lack of blemishes, they explain this as being due to the blending process and if you think about it, a more attractive face will lack blemishes and be more symmetrical.

I don't provide this as concrete evidence of what each nations view of attractive people is. I provide it as evidence that if you take a composite of many faces from a race, you will get a very attractive image. and you can see the proportions vary from a composite of white people faces to a composite of asian people faces. or black people faces, etc.

Interesting article!

Strangely....I don't look like any of the Asian women in the pictures. I most closely resemble the Puerto Rican lady.

I ran the Swiss lady picture on Haystack.ai and she got a result of 8.1. The Japanese lady got 9.3.....I think Haystack might actually be biased towards Asian women The Puerto Rican lady got 9.9!

 

[ImAllLikeOkWaitWat]

Interesting article!

Strangely....I don't look like any of the Asian women in the pictures. I most closely resemble the Puerto Rican lady.

I ran the Swiss lady picture on Haystack.ai and she got a result of 8.1. The Japanese lady got 9.3.....I think Haystack might actually be biased towards Asian women The Puerto Rican lady got 9.9!

I saw oprah get an 8 for whatever thats worth.

 

[timewerx]

Apparently the "world's best facial recognition app of 2017" thinks I'm a happy 11-year-old.

Well... You do look happy in that picture and you also look like you could be 11!

I'm beginning to have doubts whether haystack is world's best atm. I think pinkmirror.com is still more accurate (but for a different purpose in terms of letting you know how to improve your looks)

 

[timewerx]

I saw oprah get an 8 for whatever thats worth.

Anything below 9 is probably bad (not good enough)

I'll just have to keep believing the saying "beauty is in the eyes of the beholder!"

 

[Grandpa2390]

Anything below 9 is probably bad (not good enough)

I'll just have to keep believing the saying "beauty is in the eyes of the beholder!"

I'll take that. I feel like on pink mirror, 8.7ish is probably average (some of my photos get 8.7s) and then it goes up from there.

 

[timewerx]

I'll take that. I feel like on pink mirror, 8.7ish is probably average (some of my photos get 8.7s) and then it goes up from there.

Pinkmirror is most assuredly Caucasian biased. Their "perfect" standard of beauty (in their website) showed a white Caucasian female. They didn't show the male, but I assume white Caucasian as well.

Yes, 8.7 is average for a white male. I'm only 8.3 raw, 8.5 raw is the highest I got which is still below average. But if I choose female as my gender, my raw picture gets 9.2. Ironically, I do look quite feminine due to my bone structure and the proportions on my face is probably closer to the female baseline than the male.

 

[BlueAshSky]

October is here

Spoiler

 

[CodyFaith]

October is here

Spoiler

View attachment 209843

Haha that's awesome. Good job.

 

[Paulie079]

October is here

Spoiler

View attachment 209843

Love the mustache nose ring lol.

 

[BlueAshSky]

Love the mustache nose ring lol.

 

[alsughasoughaiuyfygh]

.

 

[( ͡° ͜ʖ ͡°)]

Are you going for try-outs to be the next Joker's sidekick?