322 comments

[ 2.6 ms ] story [ 248 ms ] thread
I was initially put off by the webcam permission requirement, but the terms and conditions page says it's basically an art project and they don't send any data off (unless you explicitly accept it at the end) so I gave it a chance.

I'm glad I allowed webcam permission because it was an interesting, informative, and fun look at biometric tracking.

Apparently I'm "violently average" which is not a way I would previously have described myself. According to this site the most unusual thing about me is that I read the terms and conditions before ticking the "accept" box.

Same, but I think people generally over-estimate how much different from average they are. (It's from the webcam's pov anyway) Usually I also don't like interactions with the webcam but this was interesting enough.
and now your face is linked to your ip in a db somewhere
I found that the most shocking. 96% of people who used this didn't read the conditions first.
It's skewed. I read the terms first, but then the website bugged so I had to reload. And then it said I haven't read the terms first. So....
Same, I had the wrong webcam plugged in and when I plugged the right one the website bugged so I had to reload. It said I didn't read the TOS but I did... oh well.
That's lower than I expected, tbh.
I didn't read them because I don't believe them anyway. If the website uses my camera, I'll just assume that everything is recorded and sent to advertisers and shady governments before I closed the tab. So I did not want continue with this one. It was after reading the comments here I decided to give it a go. Terms and conditions played not role.
> According to this site the most unusual thing about me is that I read the terms and conditions before ticking the "accept" box.

I’d put forward the hypothesis that people who read the terms, and who are therefore concerned about privacy, are also less likely to be willing to agree to submit the data at the end.

So you haven´t even tried without the "accept" box. It works the same, BTW.
Pretty cool to see it guess your age and BMI. Lots of interesting information. I also got the score violent average.
Morpheus: I was a prototype for Echelon IV. My instructions are to amuse visitors with information about themselves.

JC Denton: I don't see anything amusing about spying on people.

Morpheus: Human beings feel pleasure when they are watched. I have recorded their smiles as I tell them who they are.

JC Denton: Some people just don't understand the dangers of indiscriminate surveillance.

Morpheus: The need to be observed and understood was once satisfied by God. Now we can implement the same functionality with data-mining algorithms.

Sharing my face with my camera with a random website? Sure, what could go wrong... /s
Don't we all have tons of pictures of ourselves online already?
not all of us ^^
(comment deleted)
No. Zero.
Do you avoid tourist attractions and cities so that you aren't included in other people's photos? Even though I'm not an Apple customer, for example, I'm sure they could infer much of my movements from their photo database using only other people's photos (I live in a city).
Almost all, yes. The people who don't are obviously not normal, so they don't need the website.
I hope you're not diagnosing people for a living
But could be a fine statistician
And by that logic, people who do are normal so they don't need to use the website either. So no one needs to use the website :)
that's not how logic works
(comment deleted)
i will give you 1 million usd if you find 1 of mine. people are different
Worth it to be told I am more attractive than 100% of the Spice Girls.
Spice Girls back then or Spice Girls now?
It doesn't specify, so I can only assume both and which it compares to depends on how ugly the user is.
Did you read the terms of service? ;)
Do you believe the terms of service?
(comment deleted)
Yep, another magic spell akin to 'software license' that is supposed to dispell all evil.
I believe that I verified it runs locally, even offline, not sending data to a remote server.
At least for the moment
What's their incentive to change that after the HN surge?
Devil’s advocate here. What could go wrong?

If you open a website in a fresh browser context and let it use your camera, isn’t this about the same as walking down a street with CCTV cameras?

Most people have such weird and illogical views on privacy. A website collecting face pictures without anything else is pretty useless. Walking in to a retail store owned by a company using facial recognition on a huge range of owned stores is a serious privacy issue.
Face pictures and IP adresses...
And browser fingerprint and operating system.
So what's the concern here ? I'm sure I have plenty of photos of me online tagged by other people.
And the bait on the hook is not a good indicator either - an appeal to a deep drive or insecurity in many people.

Obviously not determinative in itself ,but if I wanted to harvest a lot of faces in a hurry, that's exactly the sort of bait that I would use (I can hardly think of a better one off the top of my head).

Not useless at all, there are usually enough pictures of people on the internet with names, metadata, etc attached to immediately link and identify. Basically what Clearview does (did?). I would not be surprised if this was a data collection siphon.
> I would not be surprised if this was a data collection siphon.

Except for it being sponsored by Eu - home to the most effective privacy jurisdiction on the planet.

The same organization pushing for mass surveillance?
I don't think this is a scam/data collection site (as others have recognized the researcher involved, etc.), but what would stop a random website from claiming it was "sponsored by the EU"?
web site has much more information about you that it can connect.
> fresh browser context

I get your point, but I am not sure how fresh our browser context is.

I just mean private browsing or incognito or whatever. It’s 2 seconds away in any popular browser.
> Devil’s advocate here. What could go wrong?

Could be used for a scam, by a stalker, for social engineering, or several other evil ploys. Today, there are so many possible bad actions which are happening... On average, they are unlikely, but not impossible. And who knows about the long run.

> If you open a website in a fresh browser context and let it use your camera, isn’t this about the same as walking down a street with CCTV cameras?

In my country, there are harsh regulations on public cameras. Private people are not even allowed to capture you outside from being in the background.

Could be a scam, but the guy has a Dutch accent and talks about privacy, so it's probably fine.

(I'm 54% averagely normal)

Which country is that?
A member of the European Union, there are strong regulations for all countries here.
It could use that to impersonate you in a mobile banking application through biometric authentication. It's getting pretty popular in my country (I think it's required by law or something).
You have to weigh “what could go wrong?” with “what could go right?”

The worst case scenario is far greater in magnitude than the best case scenario.

Just not worth any risk here.

A website can collect millions of faces in a few hours. A CCTV camera can only collect data from a single area.

It's more akin to those spying doorbells from Amazon and friends, which I personally would try to avoid when I can.

The concentration of data and the lack of necessity of your face being recorded in the first place change the decision making process significantly.

They give a pinky promise, that they won't do evil!
For the extra paranoid, I can confirm that it does work offline after you press the button and let it load. You can hide your face until it loads, and then cut the connection. Be sure to close the tab before you go online again.
Service Workers are a thing.

If you're extra extra paranoid, use a disposable VM in Qubes OS.

https://www.qubes-os.org/doc/how-to-use-disposables/

Or just open it in private browsing?
Wait—service workers can keep running after a tab is closed?!
Maybe or maybe not, depending on browser implementation. They can be kept in memory if the browser sees no memory need.

But they might store stuff to localStorage. Right - localStorage is also a thing. And so are cookies.

EU funded project, really not very risky.
> Government-funded project, really not very risky.

No thanks.

I'm guessing you're from the US and the parent poster is from the EU?
Correct. I don't care which government is in question, though. None are worthy of my trust or data.
I work for a company that does projects that are sometimes funded by the EU. I've never heard of the EU ever asking for something in return. You submit your project, the EU either decides to give you money, or not, and you do your project. Sometimes they check whether you've actually done what you said you'd do. No one in the EU has time to deal with the output of the billions of euro's they invest in random projects.
It's not clear if you're trying to make us feel better (the EU doesn't care about this project's data) or worse (the EU has ineffective oversight over this project -- anything could be going on).
That's not true. You clearly have no clue of the administrative side of things.
Ah thank you for enlightening us with your treasure trove of information.
If its a small project they very rarely «check» that you have done what you have done. A lot of it is self reporting. The reporting of course takes a while and you have to describes for instance your results, which wouldn’t be too hard to fake but I assume very few do this.
That's true, but there's:

A. No way for me to verify nothing shady is in the investment deal, and

B. The government may be able to turn around when the project gets big, and say "hey remember when we funded you? Yeah, now we have a say. We need you to do ___."

I don't like using software or projects funded by things I disagree with. And the government is overwhelmingly so.

(comment deleted)
It said I was way heavier than I was, way younger than I was, and way more attractive than I am. All the while I was staring into it with good lighting. Me thinks it is not useful.
It is useful when you think about that you might actually be evaluated like this in the future.
Beauty score of 5.9 out of 10. Hmm. It may be average, but that I'm still a bit peeved.
> but that I'm still a bit peeved.

You are spoiled.

I wouldnt feel too bad. It said I was 8.2, and looking at the camera view I highly doubt that, so its probably not the most accurate
I don't recall if the score went from 1 to 10 or 0 to 10 (or even if they mentioned that). If it was the latter wouldn't that put you slightly over average beauty? Even if not, it could have been worse to be determined "less than average" by a machine!
5.8 but scored 10 years younger than I actually am so I've got that going for me, which is ... I don't know.
wow, got 8. guess I'm not that ugly as I thought.
I got a 7.4, but it went up to 8 when I changed the lighting. I wonder how much is based only on the face
Almost got misgendered. I have long hair and shaved an hour ago, but still. Also, guessed I have a BMI of 20
(comment deleted)
Almost perfectly guessed my age and BMI.

Attractiveness a lot higher than i would expect (but still in range). Could be that attractiveness is related to age and BMI?

Seems im not normal when it comes to looking at dogo (why people are so sad?)

It was fun, NGL.

It was very "kind" to me, significantly underestimating my age and BMI.

I never noticed the dog picture somehow, and I guess sad is my default facial expression.

Given the usual markers of attractiveness, wouldn't you expect age and weight to be strong predictors usually?
This is fun, but it got my age very wrong, the first run through failed to guess at all, the second run through it underestimated by 13 years.

It underestimated my BMI a bit, though I am losing weight at the moment and my face does seem to be getting thinner so maybe that's thrown it, and generally it told me I'm quite attractive, so ... all good :)

I was not really prepared for being told how unattractive I am this morning.
I was a 6.3 as I stepped out of bed.

I took a shower, shaved, brushed my teeth, put on a nice shirt, fixed my hair and offered a good angle of my best smile at the camera. I was now a 9.

(comment deleted)
Interesting! I thought it would be purely based on facial symmetry.
I took off my glasses and jumped 2.5 points. If it were measuring symmetry, that wouldn't happen.
To be blunt (like the AI is): glasses are a signal of a disability, similar to crutches, or neck braces
I took off my glasses and got barely any change. Most probably GP's glasses are very very ugly.
Bald guy here -- let me assure you, symmetry is not the only factor, not for the algorithm and not for the people it represents :)
I'm also reasonably sure other factors also influence these algorithms. Hair style, framing, lighting, the clothes found on your shoulders, angle, distortion by the camera lens, and most importantly, similarity to high-scoring faces in the source data set.
I was a 6.3 but I felt pleasantly surprised because I personally think I'm a 1.5
Same. Their algorithm is broken.
I got a 4.8... the model also told me I lied about my age by nearly 20 years, so that's nice. I'm a very young looking ugly.
Same. It shaved 8 years off of me, and then called me ugly.
I’m wondering if this is scaled so 5 is quite low and 1-5 should be non-existent. Giving someone a 1 vs a “5” feels a little mean.
... I got a 5. When I smiled it bumped me up to 5.5

Fwiw, telling someone a 5 is a 1 is even meaner.

:-/

Hey, I was 5 myself, no need to take offense. We can be low self-esteem buddies.
My self-esteem is fine! A 5 is not bad : half of people are prettier than me, half are less pretty. But calling a "5" a "1" is a little mean, no? ;-) Don't really care.
Sorry to break it to you, but I'm pretty sure the wording on the site (which I don't have open anymore) suggested 5 was below the "normal" range, which presumably means it's well below the median. I'd be extremely happy to hear some more flattering interpretation.
That's actually fairly similar to how humans would be influenced by showering, shaving, etc. so in some ways that's pretty impressive.
Then you're missing the point the website makes about this and other algorithms like it being extremely unreliable. There are tons of biases in the training data, and not just ethnic/cultural ones mentioned. For starters, you're comparing a crappy webcam to a model that in all likelihood is based on people's best selfies.

The model tries to fit you into a very narrow niche. On top of that it will do so poorly.

My result: 45% Normal: "Pretty strange, in a sexy way."

Thanks, I guess? Got pretty close on age underestimated my BMI a little.

Apparently I'm just shy of being fugly and look seven years younger than I am. Which is kind of damning with faint praise, but in a dark room when I'm barely awake? I'll take it.
It guessed my age wrong, my BMI wrong, it got the thing that I came closer wrong, it got my expression wrong, and my attractiveness, which SHOULD be much higher ;)
Your username improves this comment :P

I gamed the attractiveness model for as long as I could, trying different angles and glasses on/off and hit 9.2. If dating apps use a max() across pictures that would be great.

Got my age to high, when in real life I usually get picked out as younger than I am. BMI was significantly off too. Can't argue with the not-pretty assessment though! Might give it another chance under better light later.
Wow, got my age exactly!
mine was off by 40 years .... it said I lied about my age
When I saw that it said I lied about my age (it thought I was nearly 10 years younger), I wondered whether there was going to be another metric later about whether being accused of lying made you appear angry. I was disappointed by this one omission, but greatly enjoyed it overall!
Same here, it underestimated my age by 31 years...
in v2 you will have to upload your ID to convince it
Interesting for sure. The first time I done it, I had a beauty score of 5. I then put glasses on and got better lighting, and got up to a 7.

It kept saying my age was 18-20, but I'm 32.

Finally it kept predicting my BMI as 30+, even though it's 22.5.

So it was interesting, and a pretty cool tech demonstration, but it wasn't too accurate for me.

How far is it from 'beauty score' to 'aryan score' ?
I am far from white, with black hair, black beard and brown eyes and got a 9.5 on attractiveness...

I think it's wrong though, judging by my "success" (lack of) with women.

It gave me an attractiveness of 9.5, and 5-10 years younger than I am, which has actually given me a big boost of confidence. How strange.
I've heard that line works in any number of sales situations.

And I have to say, thowawayben, that you are indeed one handsome devil. Please feel free to upvote my comment.

I think my inclination is to trust the algorithm more than any humans, which is the interesting part! It gave me a confidence boost far more than I've felt when other people have told me such things.
Interestingly, if you click on the ToS you end up on a page explaining how it works (can't link it).

The beauty scoring model was found on Github (this or this one). The models to predict age, gender and facial expression/emotion are part of FaceApiJS, which forms the backbone of this project. Do note that its developer doesn't fully divulge on which photos the models were trained. Also, FaceApiJS is bad at detecting "Asian guys".

Apparently some dating apps rates their users with these sort of algorithms. Maybe I'm living under a rock but I did not know that was a thing.

This project has received funding from the European Union’s Horizon 2020 research and innovation programme, under grant agreement No 786641.

How does a patchwork of APIs and open source pretrained models receive funding from the eu...
Indeed - I was also surprised by this. One possibility, which I hope is true, is that the authors are funded for something else (some actual research). If they spent a few hours of their workday to throw this together, they might be obligated to cite their funding agency.

I know for a fact that I've acknowledged funding agencies on papers about topics that were at best extremely tangential to my grant.

A less flattering possibility is that they wanna use EU affiliation as a badge for respect for privacy or something like that.

You have to label it clearly in your project if you have received Horizon money. It's a requirement from CORDIS. The funding is for the whole SHERPA project, not just for that page.'

The first objective of SHERPA is to "represent and visualise the ethical and human rights challenges of SIS (artificial intelligence and big data analytics) through case studies". This page does that nicely. That it is put together with available resources instead of some over engineered solution is just a plus.

> they wanna use EU affiliation as a badge for respect for privacy or something like that.

Literally my first thought.

I'm mesmerized that so many of the HN crowd are willing to let some website take pictures of them in order to present some "funny/interesting results".
There are already photos of me on the internet, which can be downloaded and analysed by anybody.
Yup. Though i am having trouble downloading what you did on the 29th of last month. Busy day?
I was spending the day dead for tax reasons.
I live in a big city where my face is captured in millions of frames by security cameras everyday.

I used android devices, now I use an iphone, and I keep cloud accounts and personal photos backups with both companies. I have linked-in, facebook, instagram and handful of other accounts with varying levels of personal information.

This cat is already out of the bag, and has been for quite some time.

I'm mesmerized that you weren't able to validate that it works completely offline and instead decided to write a comment.
"funny/interesting results" is a lot better than the deal I get from Meta.
It just goes to show you that the vast majority of ppl don't care about privacy (FWIW, I used the site as well and didn't care at all, didn't even read the disclaimer)
It underestimated my age by 15 years. Not sure if this is due to the AI being biased or simply inaccurate in general. (I'm Asian, and from the experience of me and others, Asians tend to get their age greatly underestimated in the west.)
I am 39 and it estimated my age as 10. I am not Asian.
I'm also not Asian. In late 30s and it guessed that I'm 7...
Has anyone tried sending non-face to it?
How cute is your butt? Let our 100% accurate butt inspector program calculate it for you.
Should read

"How gullible am I for uploading my face to some random data harvester?"

No data leaves the browser, feel free to check it out.