How much do I need to change my face to avoid facial recognition?
gizmodo.com
gizmodo.com
When I was in Mexico I filed a report with the airport after an employee selling timeshares was overly aggressive and grabbed my arm and try to block me from leaving. Quickly they showed me a video of my entire time with all my movements at the airport so they could pinpoint the employee.
Like the article says I think it is just a matter of time until such systems are everywhere. We are already getting normalized to it at public transportation hubs with almost 0 objections. Soon most municipalities or even private businesses will implement it and no one will care because it already happens to them at the airport, so why make a fuss about it at the grocery store or on a public sidewalk.
Other cases, like games where the user's voice changes due to excitement/stress, were incredibly bad.
This is just as interesting as it is creepy, but that's the world we live and this is hacker news. So, how quickly was was quickly. You made your report, they get the proper people involved, and then they show you the video. How much time passed before you were viewing the video?
For someone that plays with quickly assembling an edited video from a library of video content using a database full of cuepoints, this is a very interesting problem to solve. What did the final video look like? Was it an assembled video with cuts like in a spy movie with the best angles selected in sequence? Was it each of the cameras in a multi-cam like view just starting from the time they ID'd the flight you arrived on? Did they draw the boxes around you to show the system "knew" you?
I'm really curious how dystopian we actually are with the facial recognition systems like this.
It records the license plates of all cars entering and leaving the parking lots. You can associate names to faces which we do for all employees and the system automatically records when people enter and leave buildings. You can even just tell it to find all people with a blue shirt in a particular camera in a time window. It can automatically detect people shouting.
No idea how widespread it is, but in Singapore airport the system is tightly integrated. You are "tagged" when you check in, and "tagged out" as you board, with your appearance associated with your intended flight details. If you miss your flight or otherwise spend too much time in the secure zone, you are highlighted in the system and will eventually be approached. Arriving passengers are also given a time limit to take their next action, be it clear immigration or enter transit, and lingering will also trigger a response.
All in the name of safety and security but I can't help but feel a measure of discomfort with it all.
You don't have to have your photo taken to enter the US if you're a citizen, but who wants to deal with the hassle? And on and on it goes.
go ahead and decline while the cop is holding your passport.
When my spouse and I crossed through US customs this past spring, they called us by our names and waved us on before even getting our passports out to hand to the customs officer. This was at BWI, fwiw.
Immigration - which is the process of becoming a US permanent resident and/or citizen - is handled (mostly) by USCIS.
Other visas are handled by the State Department (foreign ministry).
Not an expert, but this is my understanding.
I have Global Entry but I don't think the US even has a customs form any longer.
>very few exceptions
You've got it backwards. If you're an American you're probably traveling through freeports or low tax regimes like Singapore, UK, etc. and don't realise how regressive most regimes are. In places like Hungary, Angola, SEA -- where tax can be in the range of 30-50% you will be lucky not to be shaken down by a customs agent before leaving the luggage carousel.
It's much, much more annoying in Ireland, where US immigration happens in Dublin (an affront to Irish sovereignty, but that's another matter) - so being delayed can mean missing your flight.
I’ll bite. Why do you think it’s an affront to their sovereignty? It’s entirely voluntary and it’s something the Dublin airport (and the dozens of other airports in Canada) actively seek out to get direct access to the domestic side in the US.
The US does not force any airports into these arrangements.
(The popularity of that airport for CIA torture flights also doesn't help the case, even if not directly linked)
FWIW, I recall reading that the program in Canada is reciprocal, and it is simply the case that Canada hasn't decided to operate any corresponding facility in the US.
Let's talk about Toronto or Vancouver to set aside CIA whatever. What particular convenience does it provide for the US government to do it there vs on the US side? AFAICT that would save the airline that brought a person who got denied a bit of trouble - vs having to take them back to their departure airport - but not be a particularly huge convenience or burden for either government at a higher-up level.
It reduces legal accountability (I know the US courts have generally exempted border operations from the constitution anyway, but that interpretation could change in the future) and makes it easier to prevent people from e.g. landing and claiming asylum (yes there are measures to penalise airlines and oblige them to return passengers, but they're not always fully effective). More subtly it means there's less pressure to have reasonable border rules, since turning someone away before they board is lower-stakes. And having an official, pseudo-law-enforcement presence in a country is valuable almost in itself.
Well maybe. What happens when you try to exercise that right? If it turns out that those US border officials falsely imprisoned you (under Canadian law - if the Canadian courts are even willing to hear the case), what consequences will they face?
As said policies become increasingly inhumane I think Ireland should consider removing this arrangement. But you are right, Dublin Airport themselves do benefit since it makes them more attractive, especially as a transfer airport for people going to the US from Europe.
If anything, it seems to me that the USA agreeing to perform immigration checks in Ireland and accept them when you reach the USA is a(n extremely mild) limitation to US sovereignty, not to Irish sovereignty.
rub my jeans sailor. no 3d xrays for me.
Someone got fired that day.
Case in point - one time the cash machine on the bank exterior disbursed an extra 20 to me. I counted it carefully, sure enough, an extra one. So I went around to the other side, went in, and handed it to the teller with an explanation. She about fell over in shock.
It's kinda sad, actually.
Without it, I think all the gingers would have pointed at me rather than me being tired and making a mistake.
You may be overestimating how many unique/different people travel through airports, especially more than once or twice to notice the tracking. People who travel once or twice total in their life by air, (are usually easy to spot), far more concerned with getting through a confusing hectic situation then noticing or even knowing that using facial recognition is new and not simply a special thing (because 9/11). And, the majority of Americans have travelled to zero or one country, last time I saw numbers on it. That country is usually Mexico or Canada where they drive (or walk).
I think once it starts trying to hit close to home where people have a routine and are not as stressed by a new situation and have the bandwidth to--at a minimum--take a pause, will ask questions about what is going on.
From a practical point of view to avoid getting caught if you look at the phone snatchers, they wear hoodies, balaclava type cycle masks and generic black tracksuit like clothing. If you look at the photos of the NYC shooter he slipped up in not wearing a balaclava type mask and in having distinctive clothing and backpack.
Well that’s the norm with all surveillance: it pretty much never helps you and might hurt you, regardless of the promises. Obviously after decades of constant spying, men are still getting ads intended for women and vice versa, and yet micro targeting is changing election outcomes. Banks and governments watch every single transaction, but it doesn’t reduce the administrative burden of compliance for tax paperwork. Airport experiences are worse than ever and at greater expense, but anyone with a few brain cells to rub together knows that it’s just security theater. Even more basically.. google reads all your email and searching for that exact phrase you know you read or wrote just a few weeks ago somehow turns up zero results.
This will all just get worse, because as the amount of data collected increases, everyone can be suspected of something just because of coincidence. Your insurance company is getting the memo about your poor diet or know that you’re driving too fast, and just won’t bother to find out about your healthy exercise regimen or that your job is driving an ambulance! To be presumed innocent you’ll need to opt into more data collection or disclosures of course, that’s the way it goes, but this only makes things worse because the extra data is just more stuff that can be used in a case against you.
Out of sight, out of mind. If there isn't a large video camera tracking them as they move across a shop or down the street, I'm not sure many people will even notice.
Privacy advocacy orgs should have contests for tracking people using publicaly available video feeds, or something of the sort.
Let people search their license plate to see how easy it is to track all of your movements. Maybe put up a few high res webcams in the vicinity of a legislature building for maximum effect.
Also, practically, the advocacy groups would need to get access to the surveillance feeds or deploy their own hardware - which I just don't see happening
The problem I keep running into is a real world take on the Trolley problem[0].
Do you publicly publish all data, which:
1. Reduces its sellable value
2. Makes people aware of how much they are being tracked and profiled
3. Gives back a small bit of agency over ones data by knowing where to send delete/remove request to make data brokers honour local laws
However, doing so would also:
1. Give easy access to abuse victim data, putting them in further harm
2. Give actual stalkers an easier path to their targets
3. Other harm that I can not fathom at this point in time
I don't know the answer, maybe mask the address part, or do like Strava and set a blocking geo fence around home/work addresses. For location tracking keep it months behind and remove/mask anything remotely related to health services (mental and physical).
Just like everyone though, I’m just going to gripe here and move on with my life as the mass surveillance infrastructure rollout proceeds
You're right about just trying to get through the process. I was the only one in our family who was like, "No idea why they were taking pictures of everybody when they just whisked us through customs." My Dad snapped that it was because of 9/11 and we weren't lawbreakers so just deal with it.
The comment was interesting since my Dad worked for Lockheed Martin for 30 years and used to travel constantly. He mentioned the idea of a "high trust society" is now gone forever and facial recognition and other technologies are now necessary to give that trust back to the general public so they can feel safe with air travel.
But for nation wide public search the false positive rate is just way to high for it to work well.
Once someone managed to leave a "local/time" context (e.g. known accident at known location and time) without leaving too many traces (in the US easy due to wide use of private cars everyone) the false positive rate makes such systems often practically hardly helpful.
Then again, it depends on where you are. One could have thought that finding a specific guy in NYC after you had him on camera at a given time would be easy, but they aren't so easy to locate immediately.
I'd feel much better about if we heavily surveilled the use of surveillance. E.g., every access is recorded both in terms of metadata and in terms of generating video of whoever's looking. And if I'm in something they're looking at, I get notified (barring temporary legal exceptions for open investigations and the like).
Because my business is my business and nobody else's. Full stop.
So, it stops being solely your business, and starts to become slightly others', as well.
Whose? And under what circumstances? Please be specific and include appropriate legal precedents. Thanks!
[1] https://law.justia.com/cases/federal/district-courts/new-yor...
[2] https://news.bloomberglaw.com/litigation/madison-square-gard...
[3] https://codelibrary.amlegal.com/codes/newyorkcity/latest/NYC...
Where's the case law and precedent that says your business is only your own, even when on a public sidewalk or in a grocery store? If you're going to make such unreasonable demands, can we start with your own claims, since you made them first?
Public roads should be a clear case where your business is everyone else's business since you're hurtling down the road in an increasingly heavier vehicle, but we're far from being able to acknowledge that.
I'm admittedly biased from spending a few years steeped in the cryptocurrency community, where literally everything has a hidden (or not!) self-serving agenda, however. But even beyond that realm, I see far too many privacy advocates whose examples of "reasons why you should want privacy" end up being examples of hiding bad behavior (infidelity, etc). If you couple this with anti-social people being privacy advocates out of necessity, it ends up reflecting very poorly on the privacy community as a whole.
I guess we'll see, won't we?
Being seen by others, being recorded incidentally, being recorded constantly, and being recorded and analyzed by machines in real time?
How am I supposed to take you privacy advocates seriously when you make such wild logical leaps?
Most people don't expect others to look away while you pick your nose at the grocery store. The statement about defending privacy in public is almost always about tracking and the ease of it.
Please re-read the thread, without injecting your own pre-conceived notions into it? The statement was simply this:
"Because my business is my business and nobody else's. Full stop."
When it comes to being in public or at a grocery store, this is simply untrue. Being in public involves interacting with other people, at which point it inherently ceases to be just your business, and starts to be others', too.
That entitles them to some amount of say in it, however small it might be, depending on the context.
No where did I say anything about the degree of surveillance?
Because usually, when someone says "my business is my business", they don't mean "no one has the right to look at me when I'm out in public". That kind of statement about my business being mine in public, is usually tied to tracking and/or persistent, shareable surveillance video.
If others are free to observe you while you are in public with them, are they not also free to do whatever they wish with that information? Same as you are free to attempt to share as little as possible with them. Public spaces are commons, and private spaces are subject to the rules of their owners, as you are also bound by them, in some form.
Once you leave your own property, what you do ceases to be your business alone, and begins to also be someone else's. There are clearly matters of degree, but it's also impossible to be in public without broadcasting copious amounts of personal information. If others in public are equipped to collect that information in whatever form it takes, why is that suddenly wrong?
Disagreeing with that tells me that by saying "my business is only my business, and nobody else's" you really mean "no one has the right to observe me/interact with me in public". Otherwise, you'd have to agree with my earlier statements, right?
It is not. Full stop.
Legally, it may not be. But it's creepy and invasive. Doxxing is a net negative to society. And more than individuals catching me in photos or videos incidentally, is the problem of surveillance and tracking, something I have mentioned repeatedly and you have failed to acknowledge. First came CCTV, then comes ubiquitous cheap video storage, now comes AI that can analyze that video in real time for identification and behavioral analysis.
I'll flip your question on its head: why is it the default that people should consent to ubiquitous machine-based activity monitoring by all their peers and their government, just because the technology exists? What's the benefit?
If society has lived without ubiquitous surveillance and automated behavioral analytics and tracking for all history, why is it suddenly right?
Creepy to you, maybe. It's creepy to me that some people want to live their lives as faceless, nameless ghosts in a modern society!
> Doxxing is a net negative to society.
Is it? You're making a very bold claim here and stating it as fact, as if it's somehow self-evident. It isn't, though.
> And more than individuals catching me in photos or videos incidentally, is the problem of surveillance and tracking, something I have mentioned repeatedly and you have failed to acknowledge. First came CCTV, then comes ubiquitous cheap video storage, now comes AI that can analyze that video in real time for identification and behavioral analysis.
I haven't failed to acknowledge them at all! One of my first replies in this thread explicitly acknowledged such things.
I just fail to differentiate ubiquitous surveillance from incidental surveillance. To me, the latter is simply an extension of the former. Yes, scale can make things markedly different, but it's not inherently negative. The scale of our ability to communicate has increased drastically over the past few decades, and while it has come with some bad effects, increasing the scale of our ability to communicate is not inherently bad. In fact, I think a lot of the negatives that have come with mass communication come from the relative anonymity of the mediums.
You know, the good ol' "people feel perfectly fine saying things they wouldn't say to someone's face, when they're hiding behind a keyboard, screen, and pseudonym" problem. It's not always a problem, yes, but I see far more assholes taking advantage of it to escape societal consequences than I see afflicted minorities escaping unfair judgement.
> I'll flip your question on its head: why is it the default that people should consent to ubiquitous machine-based activity monitoring by all their peers and their government, just because the technology exists? What's the benefit?
I can think of all sorts of benefits!
Better security, more accountability for anti-social behaviors, better health management, discoveries of all sorts of social phenomena that can be empirically documented, etc etc. There are vast, untapped reserves of information that can and should be used to better society as a whole, rather than discarding it all as some kind of detritus.
The vast majority of any individual's "information footprint" is utterly wasted, and you're arguing for even more wastage. I'd love for there to be preventative diagnostics for health conditions, that can do things like tell me I have early stage cancer, because it can detect information I'm emitting in some form that would otherwise go undetected. Compared to the alternative, where I only go see the doctor when I notice some ill effects, which could potentially be too late for treatment? Yeah, that seems like a net positive to me, and a far better utilization of my information footprint than simply discarding all of it.
I'd also love for it to be harder to behave anti-socially in all contexts, and to feel confident that the person I'm transacting with isn't scamming me, etc.
> If society has lived without ubiquitous surveillance and automated behavioral analytics and tracking for all history, why is it suddenly right?
Let me flip this question on it's head: if we've been living in ignorance of something for all history, when we discover that ignorance, why should we endeavor to continue in it, when there are better alternatives?
I'm not arguing that such things like "anti-social behavior" are cut and dry things, nor that my definitions of such things should be accepted by all. These are things we can and should debate, and come to agreement on in democratic ways.
But to simply be against it because "it's too hard", "I don't like it", and "we've always done it that way" strike me as a terribly ignorant approach.
>Is it? You're making a very bold claim here and stating it as fact, as if it's somehow self-evident. It isn't, though.
Okay. Please post your full name, physical address, telephone number(s), employer, spouse and childrens' names, ages and birth dates.
If it's not a negative, you should have no problem doing so, right?
Faulty logic - it can be bad for specific individuals while being a net good for society at large.
NOTE: I am not the OG claimant, I'm not supporting their argument - I'm merely pointing out your rebuttal is weak (to the point of not working at all).
How is every person living in fear that they can be identified, stalked, subjected to abuse or even killed just because some rando doesn't like what they said on some internet forum a "net good for society at large"?
Please do explain.
As for my "rebuttal" being weak, I merely called for GP to have the courage of their convictions. And since (based on their participation in this discussions) they apparently have neither, I thought I'd point that up.
Don't like my writing style? Note my username and feel free to ignore me.
That has nothing to do with your argument above re: one specific person being weak to the point of nonexistent - you've gone off on a tangent.
> Please do explain.
It's self evident to those with a background in formal debate, mathematical logic, general reasoning, etc. If you honestly can't see that and are interested in improving your comment skills in a technical forum such as HN then you may want to look into that.
> I merely called for GP to have the courage of their convictions.
Their claim that Doxxing might be a net good has little to do with it being bad for individuals doxxed, nor even being bad for every individual actually doxxed.
"Net good" and metrics for good|bad are the concepts you'd need to firm up and address here.
> Don't like my writing style?
It's the unsubstantive rhetoric in the guise of reason I addressed - that done I suspect we're done.
Perhaps with practice you'll do better in future than lurking about sniping at week old comments secure in the knowledge they likely won't even see your fluff.
I'll be sure to give them the attention they deserve!
Have a great day!
Sorry for the late reply.
Exactly.
From a fairly broad perspective, I do mean "my business is my business," but as another commenter noted, it's not only my business. And I agree. It's also the concern of those with whom I interact, both directly (e.g., talking to a cashier as I make a purchase at a grocery store) or indirectly (e.g., whether or not I litter).
In those cases, it's also the cashier and whoever maintains the street's business as well. Which is so blindingly obvious I didn't think it needed explication. My apologies for any confusion.
In any case, I find the idea that I, Nobody9999, should be tracked, surveilled and/or otherwise profiled, in the course of my everyday activities to be quite offensive.
Unfortunately, unless I want to live "off the grid" (i.e., cower in a leanto out in the woods somewhere), I have to submit to some of that. And more's the pity.
But that doesn't mean I have to like it. Nor does it mean I have to pretend it's not a direct affront to my (and everyone else's) privacy.
I'm really confused, and you're not making any sense.
Some countries have strong privacy laws, such as Switzerland.
The premise: A woman who's not well off financially after losing her job signs up for a study in which an advanced robot surgically alters her face ever so minimally so as to use her as a test case for the company's state-of-the-art/bleeding edge (sorry) facial recognition software.
She signed up because having become unemployed with no prospect of future employment, her husband's job as a gig-handyman which is mostly pest control and pays terribly, and two young children, she fears being evicted from their apartment.
The study offers a huge payment in advance, enough for their family to live in comfort for 10 months without any other income source.
One problem soon becomes apparent: in altering her appearance ever so slightly, her family and everyone she knows are taken aback: she look just like she used to, but somehow not quite: the study is intended to see how surveillance video handles faces in the uncanny valley — by creating them.
NO — I have not ruined the book if you're thinking about reading it: my introduction above happens early on, following which the story explodes in unexpected, compelling directions.
This book is beautifully written: it's sci-fi, the sixth book by a highly regarded and awarded novelist.
Read the first 19 pages (of 244) here: https://www.amazon.com/Hum-Novel-Helen-Phillips/dp/166800883...
Hopefully folks understand that this is dystopian rather than a roadmap to their next product proposal
But in those two decades, the state of facial recognition software had rapidly increased and she was recognized by a camera at a mall and matched to a national database of known criminals. At first police thought it were an error but after taking DNA evidence, it was confirmed to be the same person, and she was summarily executed.
In this day and age, I don't think anyone can truly hide from facial recognition.
Possibly the actual match level was higher. But still, the way facial recognition seems to work even now is that it provides a consistent "hash value" for a face but with a limited number of digits/information (). This be useful if you know other things about the person (IE, if you know someone is a passenger on plane X, you can very likely guess which one) but still wouldn't scale unless you want a lot of false positives and are after specific people.
Authorities seem to like to say DNA and facial recognition caught people since it implies an omniscience to these authorities (I note above someone throwing out the either wrong or meaningless "97.3% value). Certainly, these technologies do catch people but they still limited and expensive.
Why would you assume that?
Still, you're right that at those scales these systems will invariably slip once in a while and it's scary to think that this might enough to be considered a criminal, especially because people often treat these systems as infallible.
Anything is possible, but it seems from her own actions for years up until her execution that it was in fact her and she only denied it to the local police initially, hoping to be let go.
Nitpick: Summary execution means execution without due process. As per Wikipedia there was a quite thorough legal process all the way to the supreme court.
"On September 9, 2021, Lao was sentenced to death by the Nanchang Intermediate People's Court for intentional homicide, kidnapping, and robbery. She was also stripped of her political rights for life and had all of her personal property confiscated. Lao appealed her conviction in court, and the second trial was held on August 18, 2022 at Jiangxi Provincial Higher People's Court. Although Lao admitted to being an accomplice to Fa, she claimed to have only done so in fear of losing her own life, as Fa had physically and sexually abused her throughout their relationship. On November 30 of the same year, the court upheld the death sentence. On December 18, 2023, the Nanchang Intermediate People's Court carried out the execution of Lao Rongzhi, with the approval of the Supreme People's Court."
However I would nitpick that while summary executions do include those without due process, the defining characteristic is simply speed. If the execution happened uncharacteristically fast compared to typical executions, even if all due process afford to her was followed, then she was still summarily executed.
In practice a free and fair trial can't be very fast though.
(As the models get better, the produced cloaker retains its ability to fool the model, while the “minimally perceptually obvious to a human” property is what gets sacrificed — even their 2022 version of the software started to do slightly-evident things like visibly increasing the contour of a person’s nose.)
A bit naive that, it’s too late since data is already mostly available and it just takes a different government to make this protection obsolete.
That’s why we Germans/Europeans have tried to fight data collections and for protections for so long and quite hard (and probably have one of the most sophisticated policies and regulations in place) but over time it just becomes an impossibility to keep data collections as low as possible (first small exceptions for in itself very valid reasons, then more and more participants and normalization until there is no protection left…)
It's probably not too late for us, either. Facial recognition by skull shape is still a concern, but only if the bad guys get up-to-date video of us. Otherwise, all they can do is investigate our historical activity. Other types of data have greater caveats preventing them from being useful long-term, provided we not participate in the illusion that it's "impossible to put the genie back in the bottle".
Start small. Get your kid's school to take the CCTV out of the toilet rooms. There's no such problem as "facial recognition" or "mass surveillance": there are many specific instances of it. Fight those.
I think this on shopping trips routinely. I don't like being surveiled and even though I have nothing to hide (I've never shoplifted in my life!) I hate the persuasive nature of it all. I don't even mind being followed by a human that much, but I do mind algorithmic analysis that is far more effective, scary, and invasive. Sadly I think the answer to this experiment would be being asked to leave or an uncomfortable chat with a policeman. Nevertheless I silently would like someone braver than me to try it. You're allowed to wear a light on your clothes -- why not make it an IR one?
"Dazzle the camera" is an idea that sounds good when you fail to think about the whole system, and instead hyperfocus on one component.
I think you're overestimating the appetite for confrontation that many shops have.
Not, again, that confrontation is impossible, just that it's not guaranteed.
Taboo opinions inspired by W.O.P.R. Avoid playing the game:
- Stay clear of areas with cameras when possible. Revenue impacting.
- Do Zoom or Jitsi calls with businesses and associates when you can.
- Become self sufficient. Stop spending money when it is not required and have healthy groceries delivered to you. Reduce tax revenue.
- Work from home if your company permits it. Go mostly off grid.
- Hire someone to run errands for you when they can not be avoided. Pay cash to a neighbors kid to run into town.
I know none of this will be popular with anyone but I am that guy.
That said, let me just help dash any hopes of fooling government surveillance right now. Any competent Nation State that has an axe to grind with you specifically, already has you in their dragnet. They already have enough information to match your face in grainy analog B&W surveillance footage from an ancient grocery store camera. You're not beating those short of significant cosmetic surgery or prosthetics of some sort, and even then, if they want you badly enough then they'll just pull partial prints off something you touched and validate that way.
Always remember the first rule of security: if someone really wants something you have badly enough, there's nothing you can do to stop them. With that in mind, plan accordingly. It's why I don't go to protests myself, or otherwise engage with events where I know facial recognition tech is deployed: I'm in that data set, multiple times, with pristine reference materials, simply by virtue of past work (not including the updates via passport photos or Global Entry access). My safest bet is simply not to put myself in that position in the first place, and that's likely yours as well.
Fortunately that's not true of governments. Although your government may be presently compromised it is possible, via democratic processes, to get it changed back to uncompromised.
Therefore we might say, it's easier to change your government than it is to change your face. That's where you should do the work.
basically the moment you apply them to a huge population (e.g. all of US) and ignore temporal and/or local context you will find collisions
especially when you consider partial samples weather that is due to errors of the sensors used or other reasons
Innocent people have gone to prison because of courts ignoring reality (of biometric matches always just being a likelyhood of matching not ever a guarantee match).
So sunglasses and a mask then. Who cares if it’s ‘cheating’.
1. move the distance between your eyes from the center of your face a random amount
2. move both eyeballs up or down a random amount
This will defeat a vast majority of simple systems. However there are far more sophisticated ones that are slower and require more resolution:
1. mess your jaw line, cheek bones, nose bones, and depth your eyes sit inside your head
Finally the creme de le creme which even identical twins are as different as dogs and cats:
1. get the white of your eyes tatoo'd with new vasculature.
It's also actually really easy to change your gait if you want. Just watch someone, and then copy how they walk. Start by paying attention to whether the hold their more fixed "center" of movement in their chest, abdomen, or waist (or where in between), then match their degree of stiffness or sway, and you're most of the way there. It's a pretty common acting exercise.
Years ago it was announced that some Chinese cities would use gait recognition for surveillance, but I don't know if the deployment stuck. I remember a video showing off the tech, although I can't look for it right now.
That said, I'm sure it exists.
But yes, weapons are stronger too.
Things aren't black and white.
They totally can (see China, People's Republic Of), it's just in many places the authorities have not yet chosen to.
It's pure fantasy to think you'll be able to run around an resist (or even just ignore) Big Brother with a cell phone running Signal.
Computer technology has an asymmetry that, despite 90s-era propaganda, actually favors the tyrants. It's time we acknowledge that.
> But yes, weapons are stronger too.
Not in any way that's meaningful when discussing tyranny.
But moving anywhere at all by any means at all is a privilege. Driving is a privilege, walking is a privilege, flying is a privilege, biking is a privilege.
Of course electronic payment systems are a privilege, health care is a privilege, internet is a privilege, school is a privilege, jobs are a privilege.
It seems like the woman in the example is using a CV Dazzle makeup style and it doesn't fool the algorithm.
Other makeup styles work better, although I think that's probably just a short term solution (tweet is from 2018) before any new trendy makeup style is added to the training dataset.
There are multiple common cosmetic surgeries that involve eye shape.
> And now your face won’t match your driver’s license or passport, so traveling will be really difficult for you. So, honestly, why bother?
My drivers license photo went un-updated for over a decade. I didn't look remotely similar to my teenage self, and not a single person cared. Excepting one airport security person who commented on how old the photo was.
But such are the times.
The part of the UnitedHealthCare situation that seemed kind of amazing from a personal perspective was how little the suspect actually seemed to look like the public photos released.
If I were calculating "confidence", that would have been relatively far down in the percentages. Vaguely similar to a lot of euro-background ethnicity faces. Maybe there were other photos the police had and didn't release. However, from what was available, with two photos showing the mask removed and one with somebody smiling, not a great match.
Tried four reverse images searches online. Started with 1920x1280 release (blurry), and got down to two (2) ~550x550 images of the cropped face. Results from four attempts were:
1) Complained about quality, gave back electro-music looking guy in rainjacket
2) Not a valid face on both photos
3) Gave back police publicity photos as only results
4) Not a valid face on both photos
Like e.g. where I live they tested some state of the art facial recognition system on a widely used train station and applauded themself how grate it was given that the test targets where even recognized when they wore masks and capes, hats etc.
But what was not told was that the false positive rate while percentage wise small (I think <1%) with the amount of expected non-match samples was still making it hardly usable.
E.g. one of the train stations where I live has ~250.000 people passing through it every day, even just a false positive rate of 0.1% would be 250 wrong alarms, for one train station every single day. This is for a single train station. If you scale your search to more wider area you now have way higher numbers (and lets not just look at population size but also that many people might be falsely recognized many times during a single travel).
AFIK the claimed false positive rate is often in the range of 0.01%-0.1% BUT when this system are independently tested in real world context the found false positive rate is often more like 1%-10%.
So what does that mean?
It means that if you have a fixed set of video to check (e.g. close to where a accident happened around +- idk. 2h of a incident) you can use such systems to pre-filter video and then post process the results over a duration of many hours.
But if you try find a person in a nation of >300 Million who doesn't want to be found and missed the initial time frame where you can rely on them to be close by the a known location then you will be flooded by such a amount of false positives that it becomes practically not very useful.
I mean you still can have a lucky hit.
This is a very handy guide: https://en.wikipedia.org/wiki/Confusion_matrix
Where the line is always drawn for me, at a minimum, is what they do with the video and who has access to it.
Video should always be deleted when it is no longer reasonably needed. That timeline would be different for airports vs convenience stores, but I'd always expect the scale of days or weeks rather than months or years (or indefinitely).
Maybe more importantly, surveillance video should never be shared without a lawful warrant, including clear descriptions of the limits to what is needed and why it is requested.
That's a much bigger can of worms, one that reaches well beyond just airports. Many modern societies are extremely complex and assume individual access to a long list of resources and services.
Its a pretty slippery slope though. People can absolutely choose not to fly, it isn't a basic requirement for life. The slippery slope leads to larger and larger government - as long as society continues to create implied requirements on the individual it seems reasonable to give more and more power to a central authority to ensure everyone can have that access.
It sounds reasonable enough, though there isn't a good guardrail built in to avoid eventually building a totalitarian or communist state as so many things are now " basic necessities."
Not a great reference, but:
"Religious Head Coverings: A Comprehensive Guide"
<https://headcoverings-by-devorah.com/religious-head-covering...>
Wikipedia has a listing as a section of this article:
<https://en.wikipedia.org/wiki/List_of_headgear#Religious>
(Many of these cover only the top of the head or a part of the head, e.g., yarmulke, fez, kofia. Others are more comprehensive.)
The non-religious and/or cultural coverings might also be of interest.
Last time I checked there was deepface https://github.com/serengil/deepface/tree/master but it was far to work as well as that
Apple Face ID is always learning as well. If your brother opens your phone enough times with your passcode, it will eventually merge the two faces it recognizes
This will bring an end to crime and herald a massive revival of public trust and socialization.
- ubiquitous surveillance is here (your point broadly)
- the data engineering to work the data isn’t quite there, or isn’t full spectrum in the manner you argue (what prevents your theory as of now)
However, what is clunky tech today can be scaled and effective tech tomorrow, so maybe your argued future is possible, if not likely.
https://www.mitre[.]org/news-insights/publication/decipherin...
It will bring an end to poor people crime. Rich people crime will become uncontestable. Welcome to the new feudalism, same as the old.
beards are to easy to change, and masks had been very common for some time and cover more then beards
>You walk in the store and are ID'd on camera
>You buy everything and use your credit card, which is linked to your ID, since the checkout has a camera too.
>You use your email once, or your address once, with that same credit card, all connected... now they have your email and home address. You significant other has the same address? Everything is linked.
If you really want to get crazy, you can combine voting records too. Based on primary ballot numbers, you can figure out if someone voted D or R in the primary with an address.
Imagine all the stuff you can get from an email address too..
Might as well just wear a face mask and sunglasses. If your FaceID can't recognize you, neither can the other systems.
FaceID can't recognize me if I tilt my head a bit too much to one side.
Just make sure you don't how how magnets work for plausible deniability.
Yeah, it's supposed to be better than black metal makeup and even better than that early-2010's anti-detection makeup:
https://consequence.net/2019/07/juggalo-makeup-facial-recogn...
https://adam.harvey.studio/cvdazzle
Although that was around 2018-2019 so, given how quickly face recognition is evolving, I wonder if juggalo makeup works anymore. Besides, as mentioned by many of the interviewees in OP's article, there's a balance between changing or hiding your facial features and looking "suspicious" or "unnatural" which is of course context dependent. Concerts are safer to "cheat" than airports.
Wrt. public cameras which don't have such features and are much further away and aren't supper high resolution either it maybe could even somewhat work.
(Different mechanism, similar result.)
That doesn't sound like much of a defense!
Eh party contacts maybe? I use those a lot.
It is too easy to get accused of something. And you have no evidence to defend yourself. If you keep video recording of your surroundings forever, you now have evidence. AI will make searching such records practical.
There were all sorts of safe guards to make such recordings unnecessary, such as due process. But those were practically eliminated. And people no longer have basic decency!
now imagine you're the wrong religion after the regime change.
"I have nothing to hide" is a stupid argument that leads to concentration camps
https://en.m.wikipedia.org/wiki/Internment_of_Japanese_Canad...
Nazis also used census and other civil data sources. Deliberate destruction of such records in the Netherlands is one of the legacies of WWII:
<https://en.wikipedia.org/wiki/The_Holocaust_in_the_Netherlan...>
This and other legacies of 20th-century genocide are chief reasons why European attitudes toward rampant data collection and exchange are far harsher than in the United States. Though I'd argue still not nearly harsh enough.
Hostile regime will kill you anyway. But there is a long way there. And "soft hostile" may throw you into prison for 30 years, or take your your house and family. Or will not enforce punishment on crooks. All fully legally in "proper democracy".
And "wrong religion" and "leads to concentration camps" really is a stupid argument, given what is happening last year. People today are just fine with concentration camps and genecide! It is just absurd argument used to defend corrupted status quo!
If you have a "wrong religion" change it! People did that all the time.
That's the key problem. Why do you assume you'll have access to this data?
This assumes that you have access to those recordings. If you're live-logging your life via something you're wearing all day every day, maybe - but if the government decides to prosecute you for something, what are the odds that you'll be able to pull exonerating evidence out of the very system that's trying to fuck you?
Even if a system doesn't care, it's still a hassle. Case in point: https://www.independent.co.uk/news/world/americas/michigan-s...
> An African American man who was wrongly convicted of a fatal shooting in Michiganin 2011 is suing a car rental company for taking seven-years to turn over the receipt that proved his innocence, claiming that they treated him like “a poor black guy wasn’t worth their time”.
I found this article while looking for another story that's virtually identical; I believe in that one it was a gas station receipt that was the key in his case, and he ended up spending very minimal time in jail.
How many people are in jail now because they weren't able to pull this data?
I live now in a city where small exchanges are based on a default of of mistrust (e.g. locking up the tide-pods behind a glass case - it's not a meme). The only super market near (not even _in_) my food desert started random bag checks.
The modern police state requires surveillance technology, but abusive authority has flourished in any technological environment. the mafia had no problem to terrorize entire neighborhoods into omerta for example, without high technology. i'm sure there's other examples.
i don't know the right answer, but considering the extent to which anti-social and criminal attitudes are seemingly allowed to fester, while everybody else is expected to relinquish their dignity, essentially _anonymize_ themselves, it makes me less and less have a kneejerk response to the expansion of technologically supported individualization.
Surely those with larger means have a bigger impact if they're acting nefariously? And it'd be a HUGE issue for society if our rich and powerful were wrongly accused, and couldn't implement their efficiencies and expertise across the market.
Netflix has a documentary, Long Shot, on someone who was proven innocent of a murder as footage of them at a baseball game was found at the time of the murder. They had to get help finding that footage, as the police wouldn't check.
The prosecutors absolutely did not care that video footage, and phone evidence, placed him at another location, and continued to insist on his guilt. The judge eventually dismissed the charges.