Slashdot Log In
Homeland Security Department Testing "Pre-Crime" Detector
Posted by
timothy
on Tue Sep 23, 2008 11:02 AM
from the when-dowsing-meets-voight-kampff dept.
from the when-dowsing-meets-voight-kampff dept.
holy_calamity writes "New Scientist reports that the Department of Homeland Security recently tested something called Future Attribute Screening Technologies (FAST) — a battery of sensors that determine whether someone is a security threat from a distance. Sensors look at facial expressions, body heat and can measure pulse and breathing rate from a distance. In trials using 140 volunteers those told to act suspicious were detected with 'about 78% accuracy on mal-intent detection, and 80% on deception,' says a DHS spokesman."
Related Stories
[+]
Technology: Challenging The 'Unbeatable' Polygraph 101 comments
George Maschke writes "Dr. Louis Rovner, a prominent California polygraph operator, has (through PR Newswire) issued a press release titled, 'Polygraph Unbeatable, Says California Psychologist.' All too often, such publicly-made claims by those with vested interests in the perpetuation of polygraphy (a make-believe science that offers make-believe security) go unchallenged. So, I've publicly challenged Dr. Rovner to support his claim and pointed out scientific research that contradicts it, as well as the examples of several notorious spies and a serial killer who have beaten the polygraph. See, A Public Challenge to Dr. Louis I. Rovner."
[+]
Science: Sweat Ducts May Act As Antenna For Lie Detection 120 comments
Reservoir Hill writes "Researchers have discovered that human skin may contain millions of tiny "antennas" in the form of microscopic sweat ducts that may reveal a person's physical and emotional state. This discovery might eventually result in lie detectors that operate at a distance. In experiments, the team beamed electromagnetic waves with a frequency range of about 100 gigahertz at the hands of test subjects and measured the frequency of the electromagnetic waves reflecting off the subjects' skin. Initially, the experiments were carried out in contact with the subjects' hands, but even at a distance of 22 cm, researchers found a strong correlation between subjects' blood pressure and pulse rate, and the frequency response of their skin."
[+]
Lie Detector Company Threatens Critical Scientists With Suit 367 comments
An anonymous reader writes "The Swedish newspaper DN reports that the Israeli company Nemesysco has sent letters to researchers at the University of Stockholm, threatening legal action if they do not stop publishing findings (Google translation). An article called 'Charlatanry in forensic speech science: A problem to be taken seriously' was pulled by the publisher after threats of a libel lawsuit." Online translations can be a little wonky; if your Swedish is as bad as mine, this English-language article describes the situation well.
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Full
Abbreviated
Hidden
Loading... please wait.
sensors... (Score:5, Insightful)
Sensors look at facial expressions, body heat and can measure pulse and breathing rate from a distance
...And most importantly, skin colour?
Seriously, is there anything a device like this can do that's either more useful or less invasive than a human watching people walking past and profiling/screening them on what they can see?
Re:sensors... (Score:5, Insightful)
Parent
Re:sensors... (Score:5, Insightful)
Good point. A real terrorist doesn't show signs of distress, because he doesn't consider his actions immoral. He thinks killing IS the moral thing to do.
Parent
Re:sensors... (Score:5, Insightful)
The real problem with this is that the number of wrongdoers is small while the pool for false positives is high. If 5% of people have some intent that should be picked up by this, then 4% of all people with ill intent will be picked up. At the rate, then they'd have to have less than a 5% rate of false positives just to reach the point where half the people it says have ill intent actually do. What are the chances that it's going to have a false positive rate less than 5%?
And that's assuming that 1/20 people have some intent that would need to be picked up by this, while the actual rate is almost certainly smaller. Millions of people fly on airplanes every year, yet every year only a handful try something stupid. This is security theater at its finest.
Parent
Re:sensors... (Score:5, Insightful)
The real problem with this is that the number of wrongdoers is small while the pool for false positives is high. If 5% of people have some intent that should be picked up by this, then 4% of all people with ill intent will be picked up. At the rate, then they'd have to have less than a 5% rate of false positives just to reach the point where half the people it says have ill intent actually do. What are the chances that it's going to have a false positive rate less than 5%?
And that's assuming that 1/20 people have some intent that would need to be picked up by this, while the actual rate is almost certainly smaller. Millions of people fly on airplanes every year, yet every year only a handful try something stupid. This is security theater at its finest.
You've hit that on the head. About 200,000 people go through Chicago O'Hare, just that single (though large) airport, every day. And so far, zero terrorist attacks launched out of O'Hare. The odds that a person this machine flagged being an innocent is ridiculously high, even if it is has high specificity.
Also, aside from the raw statistics of the thing, there's another compounding factor that makes this even more useless*, which is it's rather simple for terrorists to game the system with dry runs.
Terrorist organizations already tend to use people not on our radar for attacks, so if they get pulled out of line on a dry-run, we won't have anything on them and it'll look like yet another false positive. Our young jihadi goes through the line with a bunch of his buddies, and everyone who gets pulled out of line doesn't go through the next time. Once you've discovered the group of people who aren't detected by the terrorist detector/profilers/crystal ball, the hot run can proceed with little fear of getting caught.
* For the stated goal, of course, not the goal of Security Theater for which a magical terrorist detector is great.
Parent
Re:sensors... (Score:5, Funny)
Fair warning, you should go trademark the phrase "magical terrorist detector" before I do.
Parent
Re:sensors... (Score:5, Insightful)
Parent
Re:sensors... (Score:5, Insightful)
Actually, the real purpose is to pull out those kids who are nervous about leaving home for the first time going to college or something. That way they can scare them into not turning into one of those dirty liberal elitist intellectuals that would dare question the authority of the system.
Because nothing turns a kid into a conservative like a bad run-in with the cops, right?
Parent
Re:sensors... (Score:5, Insightful)
Parent
Re:sensors... (Score:5, Insightful)
If the terrorists know it's a dry run, then their responses will be different - amongst other things, if they are caught, there will be no evidence or deniability.
Still, I can't see this as having a low false positive rate.
- Guy goes home to his beloved but too-oft left alone wife is nervous over the obvious.
- Gal had to much to drink last night and woke up with someone... unusual. Worried about a few things that could really change her life
- [insert various nervousness-inducing-mental-conditions-here] sufferer forgot to take his/her medicine.
- First time flier.
The list can go on.
Parent
The Paradox of the False Positive (Score:5, Insightful)
I've stolen this from Cory Doctorow
Parent
Re:The Paradox of the False Positive (Score:5, Insightful)
Out of the 10,000 people indicated as having the disease, only one did. If the purpose of the test is to find those with the disease, then it's wrong 9,999 times out of 10,000 when it reports someone has it.
Our lovely machine that is currently 78% accurate on 'mal-intent' (sic) detection is going to incorrectly tag 22 people out of every 100 as having mal-intent. With the gp's quoted figure of 200,000 people traveling through O'Hare every day, that means potentially 46,000 people a day incorrectly tagged as terrorists. Not one of them actually a terrorist, just someone caught as a false positive.
One airport. One day. 46,000 people whose lives have just been screwed over in some manner. And no guarantee that the one terrorist that might show up once every billion of people is going to be caught by the machine.
Parent
Re:sensors... (Score:4, Insightful)
That is a clear case of crimethink [wikipedia.org].
I'm calling thinkpol [wikipedia.org].
Parent
Re:sensors... (Score:5, Funny)
Unfortunately, sarcastic bitching is not the solution.
No, but it does make it a little easier to handle as the problem gets worse.
Parent
Re:sensors... (Score:5, Insightful)
Given that he published it in 1949, he can be forgiven for not foreseeing modern computers.
In terms of showing how pervasive and evil a surveillance society can be, he's still highly relevant.
Pointing out just how eerie something like an automated "future crimes" concept is hardly just sarcastic bitching -- I'm betting an awful lot of people read that summary and thought "holy crap!!", I sure as hell did. Because, the sheer idea of being detained or hassled because some computer suggested you might be stressed is nuts. It's scary to think this could give them any grounds to act on anymore more than a very cursory level -- I mean, talk about your unreasonable search, and people being told they need to get the rubber glove treatment because some computer program identified them as stressed is lunacy.
Time was when one would have through it impossible for the USA to degenerate into a place where this would be happening. Now, it's hard to think of how one would stop it. Spending billions of dollars to make all of the scary stuff in Orwell come true is frightening to some of us.
Cheers
Parent
Re:sensors... (Score:5, Insightful)
The biggest problem with this, is between 78% and 80% of people told to act suspiciously can fool the system into believing they are intending to commit crime, logically those same people should be able to act in the opposite fashion to fool the system into believing they are not, I mean really, what are they thinking the logic of their analysis represents.
Apparently excuses for legal pre-emptive arrests for unsavoury people is the new focus, much like the no fly lists. A list of politically undesirable people who will be arrested, searched, interrogated, transferred to a prison facility whilst their identities are confirmed (which I am sure will take no longer than 24 to 48 hours). All this will be done at a range of designated choke points, like train and subway stations and, maybe even toll booths.
Adjust your political alignment or you will find you, your family, your friends subject to random humiliations, violent arrests, searches including sexual groping and destruction of private property, of course your will be released and it will all be done with a masquerade of legality. I believe some journalists have already experienced exactly this type of pre-emptive arrest at the RNC convention, I don't believe they were particularly impressed with the concept.
Parent
Re:sensors... (Score:4, Insightful)
It's easy to move heat around, so any simple thermal camera can be tricked into thinking the person looks normal. This is only useful if the camera is simple. The heat has to go somewhere, so you'd see some point being much hotter than expected, but any software designed to reject absurd anomalies would reject such a point as impossible.
Facial expressions would logically require a course at an acting school or a few minutes with a bottle of latex and a blow-drier to create a fake facial skin. Criminals would not require the skill of Hollywood. They would only need to fool automatic face recognition and facial expression recognition software. At worst, they'd also need to fool low-res, low frame-rate CCTV operators at range. Most LARP groups have experience at producing very realistic face masks. Learning from them would produce someone who could (if they wanted to) be totally secure against CCTV systems. Many ethnic profilers could logically be fooled with similar methods.
As for false positives - anyone who is ill will show higher-than-normal heat, as will anyone who has gone jogging or exercising. Anyone caught in a hot car due to snarled-up roads will be hot and show an angry, hostile expression. Many in New England are permanently in a state of anger. So, in all probability, 90% of all city-dwellers and New Englanders will be classed as potential terrorists. Of course, I've always been somewhat suspect of Philadelphia cheese, but that seem to be taking the complaint a bit too far.
Parent
Re:sensors... (Score:5, Insightful)
but using this to help narrow who to watch would be what this should be used for.
I can't disagree more strongly. When the flood the false positives start coming in, they'll quickly start dismissing them. As another poster pointed out, Chicago O'Hare alone has 200,000 people go through it every day; when several thousand of them are flagged as suspicious, you can bet that security will stop caring pretty quickly.
Parent
Re:sensors... (Score:5, Funny)
I'll give that a shot...
Parent
Re:sensors... (Score:5, Insightful)
Parent
Re:sensors... (Score:5, Funny)
because they're usually bug-eyed, sweating, twitching, and frequently high
Based on that alone they would be catching a lot of nerds out on the first date too.
Parent
More bad statistics (Score:5, Insightful)
So this device was 80% successful at picking up suspicious activity from PEOPLE WHO WERE ASKED TO LOOK SUSPICIOUS.
Wow, amazing! Something any police officer who has served a couple of years would be able to do with 100% (or nearly so) accuracy.
What is missing is an assay of how many people it would flag if they were told to behave as if they were SCARED. You know... scared of being flagged for behaving abnormally, strip-searched, tortured, and never seeing their families again. Something tells me that the rate of false positives on this machine will overshadow the rate of false negatives by a very large margin.
Parent
Re:sensors... (Score:4, Insightful)
It's okay since only a few people will get hurt in the process.
Parent
Re:sensors... (Score:4, Interesting)
Sheesh! I've never seen a bunch of geeks so opposed to developing an immature technology before! Perhaps a toning down of the pessimism would be in order, and perhaps we may see some improvements in our understanding of human behaviour, and the programs built to understand it.
It's not that they oppose the development of the technology. It's that they're fed up with privacy invasions and random harassment and see this device as a means of propagating both. Even if this thing threw up 50% correct red-flags, you'd see objections.
Besides, Big Brother paranoia plays very well here - Especially when it's accurate.
Parent
Re:sensors... (Score:5, Insightful)
Sheesh! I've never seen a bunch of geeks so opposed to developing an immature technology before! Perhaps a toning down of the pessimism would be in order, and perhaps we may see some improvements in our understanding of human behaviour, and the programs built to understand it.
It isn't the idea of developing an immature technology that upsets people. It is our well-justified fear of the government deploying immature technology. I'd rather not be subjected to a public beta-test of a thoughtcrime detector.
Parent
Re:sensors... (Score:5, Insightful)
Exactly. Pair this up with the red light cameras, and you've got enough income to drive any city out of recession.
"I didn't run that red light"
"No, but you wanted to"
Parent
Re:sensors... (Score:5, Insightful)
That's precisely the point of using an automated system instead of humans, to avoid accusations of racial or ethnic profiling.
Parent
Re:sensors... (Score:4, Insightful)
So, who are the non-humans that calibrate the systems?
Parent
Err (Score:5, Insightful)
My first thought, too... (Score:5, Insightful)
All we've got is a device which can spot normal people trying to be visibly "suspicious".
Parent
Re:My first thought, too... (Score:4, Funny)
All we've got is a device which can spot normal people trying to be visibly "suspicious".
Doc Brown: Get yourself some fifties clothes.
Marty McFly: Check, Doc.
Doc Brown: Something inconspicuous!
Parent
the end of liberty (Score:5, Insightful)
You are correct. From TFA:
It is absolutely ridiculous to think that they have produced any kind of test results that would indicate a functioning system. This is government and business at its absolute worst.
Not only is DHS trying their damnedest to become big brother, they are doing it in the most incompetent way possible.
This tech will never, ever work. All it can measure is physiological attributes. Correlation is not causation. Just because some percentage of people who are intending to commit a crime have certain physiological characteristics does not mean that anyone with those characteristics is a 'pre-criminal' and should be questioned. I weep for the future.
And even if, in some far-flung scenario, it did become functional it would still be illegal. It is invasion of privacy. Our thoughts and intentions are private. They mean nothing until we act on them. Human thought is vast and unlimited, part of our nature is boiling down the infinite array of ideas we have into action in the physical world where there are consequences. Everyone has the right to think whatever they want. When they act on it, then that action enters the territory of having (potentially bad) consequences.
What this evolves into is thought control and that is the end of liberty.
Parent
Re:Err (Score:5, Insightful)
Parent
Re:Err (Score:5, Interesting)
Yes, it does sound idiotic. My reaction was: ROFLcopter at the idea that you can successfully "tell people to act suspicious". Um, if it were possible in the first place for people to notice and control the aspects of themselves that make them look suspicious, others wouldn't be suspicious of those aspects in the first place!
Think about it: people become suspicious of others based on criteria X,Y,Z because meeting X,Y,Z reveals a higher probability of intent to cause harm. But anybody trying to cause harm will suppress any *controllable* sign that they are trying to cause harm before it's too late to stop. So the only remaining criteria people use in dermining whether they'll be suspicious of someone are those that are very difficult if not impossible to control. As a bad example: someone will only look around to see if he's being watched (which looks suspicious), if he's about to do something objectionable (like picking a lock). But he can't suppress that because then he takes the chance of someone noticing him picking the lock.
A better test would be to set up a scenario like a line at the airport where the screeners have to keep out dangerous items. Then, have a few of the participants try to smuggle items through, and get a huge reward if they succeed, while the screeners get the reward if smugglers don't succeed. Then, put a time limit on, so the screeners have to be judicious about who they check, so they only check the most suspicious. Oh, and make it double-blind as much as possible. Then, the people trying to smuggle will have the same incentive structure that real smugglers have, and thus will give off all the real-world signs of planning something objectionable.
But then, that would be too much work.
Parent
"Told to act suspicious"? (Score:5, Insightful)
The summary talks about the sujects being told to act suspicious. So, if you are told to be suspicious does this make any difference from someone who is actually planning something nasty? I suppose it is difficult to find subjects who are unaware they are being observed, and yet also intent on doing something bad. Nevertheless, I'd hypothesize there might be significant, observable differences between the two groups.
Not even close (Score:5, Interesting)
Re:Not even close (Score:4, Funny)
In other words, 22% of the time it is wrong. Saying it's right 78% of the time is pure and simple market speak.
The interesting thing about this is if people started to intrinsically act suspicious, the numbers become fudged and mostly meaningless. One way this could be accomplished is by standing around handing out complimentary eye patches, telling people it is act like a pirate day.
Parent
78% isn't the number you care about (Score:4, Interesting)
Most AIDS tests are 99%+ accurate at telling you that a person with HIV actually has HIV. They're also 99% accurate at saying a person who doesn't have HIV, doesn't have HIV. Its the combination of those two facts plus "Very few people in the general population have HIV" which makes mass one-time AIDS screenings a bad idea -- you successfully pull the guy out of 100 who had HIV, then you throw in one negative bystander, and you end up adding 99% accurate + 99% accurate to get 50% accurate.
There are a heck of a lot less terrorists than 1% of the flying public.
There is a countermeasure, of course -- you use the magic machine not as a definitive test but as a screening mechanism. Know why we aggressively screen high risk groups for AIDS? Because they're high risk -- if 1 out of every 4 screenies is known to be positive (not hard to reach with some populations) then the 99%/99% math adds up to better than 95%. Better news. (You then independently run a second test before you tell anyone they're positive. Just like you wouldn't immediately shoot anybody the machine said is a terrorist -- you'd just escalate the search, like subjecting them to a patdown or asking for permission to search their bags or what have you.)
So you could use the magic machine to, say, eliminate 75, 90, 99%, whatever of the search space before you go onto whatever your next level of screening is -- the whole flying rigamarole, for example. Concentrate the same amount of resources on searching 20 people a plane instead of 400. Less hassle for the vast majority of passengers, less cursoryness to all of the examinations.
The quick here will notice that this is exactly the mechanism racial profiling works by -- we know a priori that the 3 year old black kid and the 68 year old white grandmother is not holding a bomb, ergo we move onto the 20 year old Saudi who it is merely extraordinarily improbable to be holding a bomb. That would also let you lop off a huge section of the search space off the top.
The difference between the magic machine and racial profiling is that racial profiling is politically radioactive, but the magic machine might be perceived as neutral. Whether you consider that a good or a bad thing is up to you. Hypothetically assuming that the machine achieves, oh, 80% negative readings for true negatives, many people might consider it an awfully nice thing to have 80% of the plane not have to take off their shoes or get pat down -- they could possibly get screened as non-invasively as having to answer two of those silly, routine questions.
(Of course, regardless of what we do, people will claim we're racially profiling. But that is a different issue.)
Parent
Re:Not even close (Score:5, Interesting)
Parent
Facial experessions? (Score:4, Funny)
Doesn't matter (Score:5, Insightful)
None of that matters - what's important is the false positive rate, ie. the proportion of people with no malicious intent who get flagged up. If it's as high as 1% the system will be pretty much unworkable.
Really? (Score:4, Insightful)
Isn't this a little off-base? People who are really about to commit a crime, as a rule, will be explicitly trying not to look suspicious.
Government screws private sector again. (Score:5, Funny)
Why do I even bother?
That's brilliant! (Score:5, Funny)
All you need to do now is post signs reminding any potential evil-doers to "act suspicious" and the system will work perfectly.
Fancy that, Burka's protect civil rights. (Score:5, Interesting)
If everyone was wearing a burka, then, there's no way that this system actually works. It may seem strange, but, what right does the public have to know my face?
Will be fun at the airport (Score:4, Insightful)
Absurdities (Score:5, Insightful)
We lose more people to premature death each and every year because we have no health care than we have to terrorism in the whole of the 21st century.
fear, fear, fear, be afraid, fear, fear, be afraid.
A young girl waring a proto-board with blinking LEDs could have ben shot dead because of the hysteria.
fear, fear, fear, be afraid, fear, fear, be afraid. fear, fear, fear, be afraid, fear, fear, be afraid.
You can't say we have nothing to fear, but we have a lot of real and pressing things that need to be focused upon.
fear, fear, fear, be afraid, fear, fear, be afraid. fear, fear, fear, be afraid, fear, fear, be afraid. Threat level purple.
The U.S.A. has to re-grow our spine. We have nothing to fear but fear itself. Unfortunately, the current powers that be like to rule by exploiting and enhancing the terror of terrorists.
Re:Additional Locations (Score:5, Funny)
Parent
Re:What a bunch of BS (Score:4, Informative)
Just an fyi, the accuracy number doesn't directly tell you the ratio of false negatives. It's a measure not just of how many true positives it gets (that's the sensitivity), but also of true negatives(that's the specificity), in that it should both identify the "suspicious" correctly and correctly identify the non-"suspicious".
You can't go from the accuracy directly to the specificity and sensitivity, since it's a combination of several measurements. The result, though, will be highly dependent on the prevalence of "suspicious" people in their test, which is the ratio of how often what you're trying to detect actually occurs.
I'm willing to bet that the prevalence they used in their testing is way, way higher than it would be in real life (like 1/4 to 1/2 of the test subjects were "suspicious", while in real life the odds of a random person in an airport being a terrorist is more like 1/1e6 on a bad day). So this would skew the accuracy measurement towards detecting the suspicious and understate the importance of figuring out correctly that someone is not suspicious. The problem is that when you're dealing with something very rare, even if your specificity is very high, the odds that someone you pull out of line because the machine flagged them is in fact innocent is extremely high (it's going to be over 99% chance unless this machine is -very- specific), and if your test methodology doesn't worry as much about specificity, then it's going to be even worse.
Parent
Re:Suspcious People? (Score:4, Funny)
Parent