but that is just scanning a shape externally found on a persons body. that scanner that test for homosexuality can not scan the persons brain shape and know if s/he is homosexual.
the machine is b.s. though. i am sorry. i just can not and will not believe apersons facial dementions determines sexual preference till i see reliable research duplicated many times over in doubles blind studies by creditable researchers. i dont care care if the machine is a billion dollars or an etch-a-scketch. show me proof facial demention is connected to sexual preference.
Facial recognition technology has only recently been integrated with mobile devices and computers with cameras. Laws have yet to be made to regulate how this technology should be used.
The situation is similar to how drones first reached the masses and there was no regulation on where these drones were allowed to fly. After a few years, laws were made to prohibit drones from flying in airspace at airports.
I believe that laws will be made on what A.I. & facial recognition can do in the future. However, not all countries have the same laws.
Yeah I donât know about the science there. But there does seem to be a higher incidence of some physical features like finger length ratio, so maybe.
@macy Data points on the face is how facial recognition works similar to how some movies use dots on an actors/actresses face to coincide with the animation. An example would be the movie Avatar.
Wikipedia has a good article on Facial Recognition:
yet, there is also th mask test for schitzophrenia that has been debunked too. there is utube vid with a spinning mask that say if a person can tell the front from the back s/he is schitzophrenic. i can tell the difference and so can nany other advanced pencil artist.(because we can see how the reflected light falls on its curves. this is a dead give away to know if the mask is convex or concurve. we have trained ourselves to know how the light and reflected light looks on a face and where it should be.a mushed in face or mask does not make the same patterns.) Better science is needed to connect facial dementions to homosexuality too.
my concern is not with accurate scanning. it is with people who think homosexualty can be connected to presence or absence of certain facial demintions or features. let them prove the connection.
@macy The article stated this:
The research, which went viral this week, used a sample of online dating photos, limited only to white users, to demonstrate that an algorithm could correctly distinguish between gay and straight men 81% of the time and 74% for women, suggesting machines can potentially have much better âgaydarâ than humans.
The face mask test is a bit different than measuring a finger ratio or facial features, mainly because you can train your brain but you canât really alter your physical features.
But yes, a single study is not proof, other researchers have to be able to replicate it.
the font directly below the headline says âaccuratelyâ. i guess their idea of accuracy is different than mine. let us hope no ones life depends on that. and i hope that brand of accuracy never finds its way to a medical lab.
on the lighter side, i am not even gay. why am here advocating for their wants is a mystery to myself??? but i will read your info
age, injury, illness, surgery will alter facial featues greatly. this news is immature as the study. was it another poster who said it was tested only on males with in one race? what about others? mixed race? their study os very limited.
Plastic surgery and traumatic facial injury or scarring illness aside - the prevalence of those things is low, wouldnât substantially change overall results.
Age does change facial structure somewhat, but thatâs a factor you could potentially control for by comparing age matched subjects, or perhaps in other ways.
80% accuracy is actually not unusual in medical tests, but theyâre screening tests. The radiologist might look at your scan and say âThis is looks like cancer.â But they determine that for certain after they go in for a biopsy. AI is increasing doctors diagnostic accuracy in this area though.
Thatâs horrible, people deserve their constitutional privacy!
Not to derail the thread, but this reminded me of a job application to Target I completed the other day (did not get job) and was stumped by the personality tests they put me through, such as true or false, your coworkers can guess your mood.
I freaking hate HATE personality testing, it gives away my sza. Itâs discrimination, I think, and shouldnât be a part of a pre-interview application, thatâs wayy too invasive (the interview can be used to test for âmental stabilityâ anyways).
Back to the op, thatâs dangerous for people, not just their physical lives, but their earning potential as well; discrimination against lgbtq people is still a problem in the workforce, this AI makes them more vulnerable to discrimination in the workforce, housing market, etcâŚhopefully some hacker figures out how to trick the AI, much like the hackers figured out how to hack self-driving cars by sticking little stickers on stop signs.
weight gain/loss is not uncommon at all and greatly changes the shape of the fullness of the face and fat fillsin the creases in older in dividuals faces to give them a more youth appearance. the noses and ears continue to grow (veryslower rate) on older people.and this makes them look as if their ears have âdroppedâ. the eyes lids start their journy to be more hooded as early as age 25. loss of teeth or addtion of dentures is also very common and greatly changes the shape even on mid aged people. there is so much variation! how arevthey going to account for all that?
There are some things that age doesnât change, like the distance between your eyes. But that aside, as long as youâre comparing like to like, it should be possible. So if youâre comparing 50 year old gay and 50 year old straight men, age ought not be an issue. Likewise if you are comparing 18 year old gay men and 18 year old straight men. So, if you could develop a set of characteristics a machine could measure, and programmed them on faces of every age range, you would have the machine compare like ages and races to each other and not to other groups.
Dont get me wrong. i do understand comparing like to like. I just think the developers can not comprehend the vast amount of genetic variation of the worlds population.Not even mentioning the mixed race people.
I dunno, AI can âseeâ and make note of little differences over vast amounts of data that we canât. So I really think the question is, is being gay associated with subtle variations in facial structure. If it is, then AI can probably be programmed to detect it.
Iâm gay, and thereâs a lot of times I canât tell whether someone is gay just by looking at them, I doubt a machine could accurately tell all the time.
Itâs hard to compete with a machine, if it has a database of thousands of faces to draw from, has perfect recollection, and can measure (and remember) each and every facial feature of each of those thousands of faces.
I mean, as a person I have the ability to recognize faces. But I canât remember everybody and I donât have perfect recall. So I canât compare my ability to recognize faces to that of a machineâs. But, 20 years ago that was not the case, people were better at facial recognition.