Predicting sexuality with face recognition. No ethical approval, no consent of the users whose faces were used?https://osf.io/zn79k/
-
-
I could get 90% by predicting that everyone is straight.
-
Yeponomics
-
Although actually I don't believe that 10% of people are LGBT. It's higher. That's from a single flawed study if I'm not mistaken.
-
Yeah. Kinsey I believe. Most people say 2.5-5% though https://www.theguardian.com/technology/2017/jul/09/everybody-lies-how-google-reveals-darkest-secrets-seth-stephens-davidowitz …
-
I'm definitely LGBT. So 100% of people in this room.

-
Are we doing a
headcount now? 
-
Fuck yeah!

-
Bring out your queerness y'all.

- 1 more reply
New conversation -
-
-
50% is chance for guessing if somebody is a (cis) man or woman as the underlying distribution is 50/50.
-
Unless I've really misunderstood what you're trying to say...
End of conversation
New conversation -
-
-
The data was balanced, so I think 57% is not much better than chance. That was using their "facial femininity" feature only.
-
If I were to pick out a flaw, it would be from doing 20-fold cv on a dataset of faces where total images outnumbers number of people.
-
Ah, I missed that they had equal numbers of "gay" and "not gay" people. Nice catch!
-
And then made to decide which of two pictures was which, right? So even easier in a way?
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.