Facial recognition systems vs. 1st Ammendment Protection
I just got back from a trip to a new potential customer for my company. Of late we have been working hard at perfecting software that will be used to control fully autonomous vehicles in strip mining operations. Because of our efforts, we received a call from a company that has been awarded a contract to produce armored vehicles to a state police force in a country along the Pacific rim.
Most of the equipment in the specification is to be off the shelf fixed and trainable cameras integrated with recording and WLAN capabilities, GPS Ect.. One of the specifications though calls for computerized facial recognition. In doing some preliminary research on the current state of the art, computerized systems are 97.5% accurate compared to 97.7% human accuracy when viewing fixed (mug shot) photos. The current accuracy for what is called "In the wild" facial recognition is currently down at 33% for automated systems. This level is improving all the time, but is still limited by computational speed. I found it ironic that while there is no standard database used in facial recognition by any known law enforcement entity, many of them have been able to use photos people freely put on social media sites to generate a way of identifying people that law enforcement takes photos of. (Still think all those photos you put on Facebook are a good idea?) It is my understanding that the FBI facial recognition system has 52 million international faces, as well as one-third of Americans. (I know they have mine from the military, and the DMV.)
NOW TO THE IMPORTANT DISCUSSION....
You, I, in short everyone has the right to take photographs in public. Facial recognition can be applied immediately, or days later, or months later. If someone takes a photograph in public, and wants to apply facial recognition, do they need consent to do so? The facial recognition databases generated from photos people freely make public on social media sites, are these to be considered "consent"?
Should a company need consent from a stranger on the street before using facial recognition on them?
Some examples....
- Would a store need to get opt-in consent from a shoplifter before using facial-recognition technology?
- Should police get opt-in consent from a missing child before using this technology to find them?
- Should we have to get opt-in consent from every friend and family member before we tag him or her in our own photos?
- Do terror suspects need to sign off on being put in a database?
This is just the tip of the iceberg with respect to biometric data collection and use. Police have been using fingerprint comparison software for decades now. By connecting the dots one can see that we will get to the point where even our DNA is going to end up being a target for identification software. One hundred years from now we will have the answer to these things, but in the here in now....
Does running an "In the wild" facial recognition program fly in the face of the First Amendment when it used on a comparative photo we freely put on the internet?
Tags
Who is online
84 visitors

I'm in agreement with you NWM...... we are by our own hand are forcing change in what is going to be determined as to what is public and private. Our Constitution is a living document. I see the immovable object of the right to privacy coming up against the irresistible force of technological change being inevitable. Something is going to have to give.
This is one of the areas we disagree my friend. The Constitution is not a living document as much as it is a contract between the government and the people. And both parties are bound by it's specific provisions.
It's the ideals embodied within it's provisions that are important and the changes in language and word meanings do not change those ideals.
Technology will force changes in how those ideal are applied. same with philosophical attitudes. But the ideals are what is important and should never change.
The argument is always about how such ideals are applied, not what the ideals are. What scares me the most are those people that would throw those carefully crafted ideals out in the desire for some short term benefit.
I will agree to the extent that how such is applied will change, but the basis of our society should never change.
I only hope we don't wind up fighting another war over it.
Hmmmmmm...... Interesting. I do agree that the Constitution is the contract to which you speak, but if the Constitution is not a "living document" why did the writers provide provisions to amend and make changes to it? Furthermore, every time the SCOTUS makes a ruling, it is doing so by taking into account contemporary norms in American society against those of what is believed to have been the framers original intent. This is why opinions are rendered.
I am anything but a lawyer here, so my interpretation of opinions in constitutional law is on very minimal ground. I think we are in agreement here, but it is the semantics which are getting in the way. (Isn't that what usually happens?)
Probably are on the semantics issue for the most part. but there are subtle differences.
......every time the SCOTUS makes a ruling, it is doing so by taking into account contemporary norms in American society against those of what is believed to have been the framers original intent.
In actuality, it depends on the specific case and if the issue had been heard and decided before hand. (they do not like to duplicate their efforts) If the case has a history of previous decisions tied to it then yes the specifics of the case are matched with the reasoning of the previous cases and an opinion is rendered. (fitting the new into the old, a shortcut but an effective one. it is the whole basis of legal precedence) A case in which the issue has never been decided before goes an entirely different way. Called a case of first impression, this would be a case that the issues have to fit within the framework of the constitutions specific provisions, (including the bill of rights which are part of the constitution as written)
In this case they have to go back and frame the issue in the light of what was written about it in the day the constitution was written to determine the way the issue should be viewed legally. Then apply the founders intent from their writings as a check to conformity with what was meant by the provisions the issue falls under. Such meaning has to be the meanings the writers actually meant, not what is interpreted by today's meanings.
There is a basis in contract law that all contracts are bound by it's specific terms, in the language it was written in, AND at the time it was written and agreed to. The Constitution is written in British Colonial English. which is quite a bit different than modern American English. So because of the doctrine of laws, they have to go back to what was written with the meaning understood at the time it was written. To comply with well established (600 year old) legal doctrine.
Heller was the last case so decided.
This also means they cannot apply modern societal standards to issues before them, particularly issues of first appearance. Modern standards need to fit within the existing framework. Doing such would completely destroy the legal standards of adjudication. (I'm probably not explaining it well) And any valid basis of law. (the law could be what anyone thinks it should be)
Otherwise, I think we are in agreement?
Oh yes, almost forgot, all contracts have provisions to alter their terms. (whether specifically stated or not) AGREEMENT OF THE PARTIES. (and the constitution specifically states who the parties are and what the terms for changing it are)
Disagreements are what they have courts for.
This goes back to the age old question of security vs privacy. Terrorists and criminals are making use of new technology to do their evils. We task our police and military forces with protecting us. For the latter to succeed against forces becoming ever more clever in hiding their efforts towards crimes, what tools are we going to allow those charged with our protection to take advantage of?
I already know how this discussion is going to go but I thought this real world activity might be one for good discussion.
Some points made by NT in the past on various discussions that align with privacy vs security issue:
We see the need for ever more precision in drone strikes, and agree it was a good idea when we get the bad guy, but we have a problem with summary execution when we put it in perspective of "due process".
NSC computer systems that are designed to try to locate domestic terrorists through calls and internet usage, but require the collection and analysis of ALL calls and internet communications by the public without their consent to be made useful was found by many to be in invasion of privacy.
I have no problem with facial recognition. A few years back, there was a show on TV called "Lie to Me". It was based on a real team of people who not only had complete facial recognition (a rare type of total recall), but were also trained in the science of micro expressions. The FBI used them on a regular basis and they had nearly well over a 90% success rate. If people can do this, without permission, why can't a computer?
An invasion of privacy would be cameras in our homes or other places where there is an expectation of privacy. Out on the street, you are free game. Are you going to get upset when a guy you hated from high school stops you on the street, because he recognizes you? That is what these cameras are like.
Now, on the other hand, since they are not 100% accurate, the question is the legal status in a court of law. Remember it didn't end well in 'The Minority Report". It should only be used as a tool and not a final decision.
All fair points Perrie........
How about the generation of the data bases using social and professional systems like Facebook, and Linked In?
Hi Fly,
Hey you put it out there, it's fair game. This site is no different. We are monitored by the government, just like every other site on the internet. But this does bring up a good point. Everyone should be wise about what they put out there. Becuase once it is out there, it's out there forever. I think about that every time I post or send an email. There is no real privacy anymore. There never really was. Technology just made it easier for them to snoop.
(Oh and I should proof read what I post... I just made a few corrections on my last one, LOL!)
There you go....
The most dangerous four letter word in the English language.... S E N D!
Just gonna put this out there, a quick search for methods to defeat facial recognition software comes up with some good, and not so good, suggestions.
1. wear a mask. Hey, it worked for the lone ranger.
2. Wear a ball cap or something with a bill.
3. Tilt your head down slightly, as though you're looking at your watch or phone. Sounds eminently doable.
4. A thick, bushy beard conceals the length of the jawline and shape of the mouth and the proportions between them. That would work for me but it might look silly on some of the ladies.
5. Sunglasses. They conceal the color and spacing of the eyes.
6. Noseplugs, to change the shape of the nose. I suspect breathe rite strips would do as well.
There are some more technical tricks and some basic social engineering ones. Like DON'T POST HUNDREDS OF PICTURES OF YOURSELF OR YOUR FAMILY ON SOCIAL MEDIA. That one seems obvious to me but it probably isn't as obvious to young people who literally grew up sharing every moment.
This just scratches the surface. Do the search yourself. It's some interesting reading.
LOL good one Wheel!
LOL perfect Wheel.
Another way is to be a ''shape shifter'', which I am. It plays hell with it.
All good points..... Even without all that, current "In the wild" systems are no better than 1-in-3 for gaining a "hit". Smiling more should extend that range greatly.....