Hacker News Reader

Comment by mabbo | Hacker News Reader

I'm inclined to believe that an ML algorithm can do what the authors claim- but maybe only for the data set they're using.

The inputs are self-submitted photos to a dating website. Can we be certain that how people pose, smile, look, etc, isn't influenced by their intention- to find a mate of a specific gender? What if those same people were each asked to submit a second photo when they're told to 'act gay/straight' in order to fool the algorithm?

It could be that this ML algorithm is picking up on very different cues than intended. But I wouldn't discourage the researchers over it- try it again with a better data set, one built for this purpose, and discover just what the algorithm is picking up.

View on HN

This particular algorithm is being talked about heavily in the gayer parts of the internet, and I keep telling everyone precisely this. This is interesting in as far as it shows that, in the context of trying to attract a partner, gay and straight people present themselves in quantifiably different ways, which I suspect most people sufficiently familiar with app dating culture would have already known.

Again, I hate to keep parroting what another HN said, but I hate to see misinformation repeated:

The paper states that they used an existing net, VGG-Face, to exact facial features: https://osf.io/fk3xr/ (page 13) >VGG-Face aims at representing a given face as a vector of scores that are as unaffected as possible by facial expression, background, lighting, head orientation, image properties such as brightness or contrast, and other factors that can vary across different images of the same person. All the heavy lifting for this paper was done by logistic regression on the facial features reported by VGG-Face, they didn't train a DNN specifically for identifying sexuality. The algo wouldn't have seen clothing at all.

Basically, it's not as simple as "All the gay participants had rainbow flags in the background of their photo so let's ignore this study"

This is very interesting thanks for sharing, but I am not sure if it totally kills my point. I assumed face was isolated from clothing/hairstyle/background already, but believed facial expression (etc) was the main thing being looked at. If this VGG-Face tool is sufficiently good at making expression (etc) not be a factor, this is potentially a huge deal and some of the worries in the gay community may be justified. Excuse my generic HN skepticism but I will need to look into how legit VGG-Face is before I start thinking about what this means.

I don't think it does kill your point. The set of images used in the sample are obviously biased. They are by people posting pictures they hope to use to attract others of a certain gender (or genders). Even if it is specifically taking dimensional measurements surely some facial expressions bias the dimensions. This could still be the equivalent of a duck-face detector.

> Basically, it's not as simple as "All the gay participants had rainbow flags in the background of their photo so let's ignore this study"

Well, you've made me laugh and made an excellent point.

I suppose I should read the paper more in-depth.

I don't have a working "gaydar" but I do believe that it's real (meaning there are people who can predict better than chance). In the same way people can learn to sex chicks based on intuition after some training: http://www.businessinsider.com/the-incredible-intuition-of-p...

Is it that doubtful that a ML algorithm can reach/exceed human-levels on this "gaydar" problem?

Reply on news.ycombinator.com