> With this system, the researchers claim that the way a person walks and analysis of that individual’s footsteps could be used as a biometric at airport security instead of fingerprinting and eye-scanning, providing a non-intrusive method of identity verification.
This is one of those modalities (gait, keystrokes, tone of voice) that gets rediscovered every few years, shown to be waaay worse than e.g. iris recognition and forgotten until somebody else picks it up again.
Back when I was doing my thesis(liveness detection based on pupil dilation) there existed a system that could scan multiple irises from a distance of up to 6 metres. I assume these kind of systems have only gotten better since then.
The head of our department was rightfully sceptical as well.
Here's how they did it: they had a few decent cameras zoomed and focused so that their fields of view formed a skewed line crossing the path the subjects walked.
You only need a 300x150 image of the whole eye to perform decent iris recognition, so an eye covering 10% of a given camera's field of view was usually large enough.
But that was in 2012 - I imagine they've made some improvements.
I still think we can only survive as societies with truly radical openness - stick up all the cameras younwant, but it is legally obliged to publish everything you are monitoring - everyone gets to see what everyone else sees.
i honestly don't see any alternatives, despite the enormous implications
Publish it or you will be liable under the law for upto 4% of your global corporate turnover. For each camera.
Seems to be working so far as a threat.
Look this is a new world, just as the one where "sending children to work down mines and in factories is bad" or "dumping pollutants in rivers is bad". The laws took time to exist and time to get teeth and time to get enforced - but they feel natural now.
With good optics and fast tracking, I don't see why not. Even if your video is lower-resolution, you could use a video feed for movement tracking to direct a high-res still camera with decent optics and have a high-resolution capture of the face and eyes, maybe enough to distinguish iris details. 6 metres doesn't look hard for my off-the-shelf optical-zoom camera.
If you are rich and can afford custom optics and a large-ish array of sensors (think of Kepler's camera back, more rectangular and much smaller), you could capture different sensors independently to get different sections of a larger panorama separately, identify which ones hold high-enough res irises and pipe them through your recognition plumbing. Knowing all irises are within a certain range of heights make the problem a lot simpler WRT positioning the camera, the optics and the sensor array.
So wearing shades indoors will be illegal. It makes me wonder about contact lenses with an iris pattern printed on them? Maybe the system will flag if a persons eyes never dialate at all?
My thesis was exactly about this. We even had a lab course in which we used paper printouts of irises with holes poked in the center(so that light reflections on the cornea were captured) - that's how easy fooling those systems used to be.
The iris is constantly slightly changing its size, so with such contact lenses you can expect to be suspect no 1.
Silly walks may be freedom. I find this mostly daunting. I wouldnt mind if the technology was developed to improve our understanding of city commute, but this gets into some freaky stuff. When were all much older, I can imagine getting pulled aside by an officer because my walk is disturbing the computers.
It's relieving to see basically all the comments here expressing how dystopian this is, and my opinion is no different.
As I read the article I keep asking myself what motivates these people to do such research --- they're basically contributing to the erosion of personal freedom and privacy. It's scary.
> As I read the article I keep asking myself what motivates these people to do such research --- they're basically contributing to the erosion of personal freedom and privacy. It's scary.
I think it's better that it's public research rather than some private R&D. Public knowledge never hurts in my opinion.
Airport security is a big sham in my opinion. All this constant step-up of surveillance tech and there's no statistically significant risk to be mitigated. Terrorist attacks still happen with no real consequences.
It's just a palatable proving-ground for total social surveillance tech. That's what the real benefit here is.
Many people have a legitimate belief that security is preferable to freedom.
To the folks downvoting this objective fact: I suggest you take a look at your ability to rationally discuss things that upset you before spending much more time on the internet.
It's certainly a fact that some people prefer security to freedom. In addition freedom is a very abstract concept, and as a consequence it is much more often fetishized than adequately understood. Absolute security and absolute freedom are both essentially death.
That all gets very philosophical though, and most people can recognize a good balance of freedom and are watchful of security paranoia stepping up and curtailing that balance.
But here's the problem. This isn't actually security. Security is the straw man. This is power masquerading as security. Nobody is becoming significantly safer thanks to these overreaches of surveillance in any way that really matters. The likelihood of falling victim to a failure of airport security is outrageously low, and these security firms are taking advantage of our inability to accurately assess risk when it comes to things like this.
Absolutely, the lizard part of our brains are wired that way. Typically, folks manipulate our assessment of risk, as we are terrible at assessing that.
Most of the most awful atrocities in history fit that pattern, from Ceasar in Gaul forward.
I prefer to inject a little randomness into my attacks on gaitrecognition: I put a handful of gravel into each shoe. Cheap and effective, and no two steps are the same. Plus you get a great reflexology foot massage in the process (I kid. Reflexology is about as scientifically useful as gaitrecognition).
So now it just tries to keep track of who's where and when. If someone leaves by the school-gates during classes, their gait is checked to see if it kinda-sorta matches any student gait and if it does, whoop-whoop-whoop, ring the alarm!
Chavez High is ringed with gravel walkways. I like to keep a couple handsful of rocks in my shoulder-bag, just in case. I silently passed Darryl ten or fifteen pointy little bastards and we both loaded our shoes.
- Cory Doctorrow, Little Brother: https://craphound.com/littlebrother/Cory_Doctorow_-_Little_Brother.pdf
Unless you need a captcha in an airport, I wouldn't say this qualifies as useful... 0.7% error rate on biometrics means that of the 1000 people who will use the airport on a particular day, the system will fail 70 times, that's huge. Plus this is in lab conditions. Add a few years to the sensors, and your error rate goes through the roof.
Plus I'm pretty sure you could develop "adversorial shoes" for this...
I'm no fan of the TSA, but you're comparing apples to handguns-in-handbags here; their inability to find planted explosives in my luggage isn't the same as a failure to correctly identify that I have the right documentation to get on the plane.
You're right; failing to detect weapons (and thus letting people die) is a fatal error that can't be corrected. Failing to correctly id you is something that can be corrected by a human afterwards. I'm much happier to let a system incorrectly id someone than to let them onto a plane with a weapon.
That said, this seems like just one of many potential id methods that would comprise a larger system (facial recognition, fingerprinting, etc).
It'll fail .7% of the time for each read. You must assume you will be read multiple times at multiple locations under different lighting and motion tracked the rest of the time. Once you get a couple consistent identifications, the system can be pretty sure it's you.
With only 127 people being tested and a false positive rate of 0.7% I'd be hard pressed to see this as a viable system in its current form.
It's certainly less invasive but you have to ask how effective it can possibly be when it's known about. It's a lot harder to fake your fingerprints and retina than to walk at a different speed.
If you were a highly funded individual you could also quite easily acquire one of these systems and practice to fool it.
I don't see how this could be used as a effective security system but maybe someone could enlighten me.
In many security settings you would have to do more than train to not be recognizable as yourself, you'd have to train to be recognized as someone else.
Even a busy airport mostly already has information about who should be where.
If you combine this with many other types of sensor data you can get a pretty high dimensional array of stuff that can pretty easily form person-unique clusters. Like combining screen res, Touch Support, User agent, Time Zone, Hash of canvas fingerprint, etc [0] gets you to a pretty unique signature even though any single bit of info is essentially worthless to this end.
You could probably trick that system by dragging something heavy, deviating from your regular gait. Now where to get something heavy to carry around at an airport, I have no idea.
>>> With this system, the researchers claim that the way a person walks and analysis of that individual’s footsteps could be used as a biometric at airport security instead of fingerprinting and eye-scanning, providing a non-intrusive method of identity verification.
Lol. Does anyone here think this is possible given the realities of airport "walking". Pressure patterns? You mean the pressure my feet place on the ground as I shuffle because my laces are untied approaching security? Or my stride length as I drag my roll-aboard with its one broken wheel? Or my pace as I run between flights, limping because one foot is still asleep after eight ours in an economy-class cage.
And ... I don't drink but many people do drink at airports. Everything with your walk changes after a beer or two.
And women in heals.
>>> the process is non-intrusive for the individual
Yes... until I get pulled over for questioning because my walking pattern doesn't match my last walking pattern. TSA is already security theater. If our missteps will result in bad reviews, that theatre is now a security ballet. If you want to make your flight, remember to hit your mark.
He didn't mean literally. But these people are seriously nut. They want to scare people, that is all. We all know that the 9/11 terrorists where on the Saudis payroll. And Saudia Arabia is an american ally!
Oh course. Asymmetric warfare is usually successful.
We’re trying to swat flies with World War 2 armies. The US has been in a war state for so long there is no actual condition that will result in victory, and we have undermined the institutions of civil government with no tangible benefit.
This'll work great until someone has an event that changes their gait -- sprained ankle, different shoes, stone or wrinkled sock in one shoe, carrying luggage asymmetrically, etc., etc., etc..
They even mention this sort of problem near the end of the article, where they mention it would be a good way to pickup cognitive or physical decline. This controlled use would be a much better use of the technology, but grab fewer headlines.
This is true, but if you get enough people doing it, the anomalies no longer stand out and you start struggling to pick out the signals above the noise floor.
We'd need to make it a popular choice, the easy option in clothing/fashion/footwear.
Unless it's adopted by a significant percentage of the population, it's only a tool to draw added scrutiny. Much like when I tried to board a plane with an excessively comfortable sweater and was patted down.
Security? Whose official vehicles are in the basement parking lot?
According to the register, a Mr. Nakamura from the Treaties Bureau and a Dr. Willis, sir.
Get me the video records of their entrance.
Again. In infrared this time.
One thousand, one; one thousand, two; one thousand, three.
Now get the records of the pressure sensors for basement parking stalls B8 and B7.
Major, this is Togusa, Code 09.
What's up?
Does that Nakamura guy have a special-order cyborg body?
No, there aren't any cyborgs in Section 6. They don't even ship them to serve overseas because they're too hard to maintain. There's diplomatic considerations too.
Then, even if the other guy that was with him is a cyborg, the two of them together couldn't weigh over 1,000 pounds.
Ah, you checked the pressure sensors in the parking structure, didn't you?
Right, but neither man looks like the type to drive himself. The security cameras show only two people. But remember how sensitive the sensors on the entrance doors are? They took a full three seconds longer than usual to close after the two went in. And it's illegal to use therm-optics inside government facilities.
It's a serious violation of the National Security Act. Section 6 must be up to something. Are you ready?
With my trusty Matever, anytime.
- Ghost in the Shell 1995: https://en.wikipedia.org/wiki/Ghost_in_the_Shell_(1995_film)
This movie is full of nice quotes regarding tech and it's from 1995! Japanese people really have a fundamentally different view on tech, tech is always a tool. We Westerners love the classic Terminator plot, the classic Asimovs laws gone bad plot (I, Robot). The Japanese use tech, they sit in battle robots, they use it connect their consciousness directly to the internet. I love it.
Ok, some more from Ghost in the Shell 2: Innocence, fully off topic but hey I'm getting pretty fired up, I'm going to watch both movies again really soon... The movies are so timeless because they don't focus on exact tech, they focus on the philosophical implications of the tech, it keeps them highly relevant.
Scene 9: Golden Sky
(In a vehicle flying across the remains of a city)
Batou: "This is one of the largest cities in the Kureal Islands, it was once prosperous, originally built to be one of the most important information centers in the Far East. See those towers? This city was really something. But its dubious sovereignty has made it the ideal haven for multi-nationals cooperations and the criminal elements that feed off their spoils. It's now a lawless zone, beyond the reach of the e-police. Reminds me of the line, 'What the body creates, is as much an expression of DNA as the body itself.'"
Togusa: "Doesn't that apply to beaver dams and spider webs as well?"
Batou: "If the essence of life is information that is carried in DNA, then society and civilisation are just collossal memory storage systems, and a metropolis is simply a sprawling external memory."
Togusa: "As it says in the bible. 'How great is the sum of all thy thoughts. If I should count them, they are more in number than the sand.'"
Batou: "Psalms 139, Old Testament. The way you spontaneous spout these ancient exotic references, I'd say your own external memory's pretty twisted."
Togusa: "Look who's talking."
Speaker: "Sorry to interrupt, but we're about to fly over Locus Solus. I'm changing the route just for you, so take a good look."
Speaker: "So much for the free tour. Now prepare for landing."
Togusa: "'His legions, angel forms, who lay entranced / Thick as autumnal leaves that strew the brooks'"
Batou: "Now you're quoting Milton, but we are not Satan."
You're kidding yourself if you think the Japanese are above a good old-fashioned AI gone rogue plot. Exhibit A: Macross Plus (and the AI is an idol singer to boot). Exhibit B: GitS itself, with the Puppetmaster.
As for "Asimov's Three Laws gone wrong", Asimov himself created the Three Laws because he thought Frankenstein had been retold too many times and wanted to write stories about robots blending into human society rather than attempting to overthrow it (and imagined interesting new failure modes of these robots in the process!).
> Japanese people really have a fundamentally different view on tech, tech is always a tool. We Westerners love the classic Terminator plot, the classic Asimovs laws gone bad plot (I, Robot). The Japanese use* tech, they sit in battle robots, they use it connect their consciousness directly to the internet. I love it.*
I think your conclusion is correct in the sense that the Japanese use tech enthusiastically, but very wrong in the assumption that the reason is that they just view it as a tool. If anything, they seem to anthropomorphize a lot more than Westerners. I once attended a lecture by a Japanese professor researching robotics who basically seemed to live by the idea that if his robots looked and appeared to behave human enough, they would magically "be" human. It made it look like the anthropomorphization of robots its perfectly normal even on academic levels over there. Also, note how until Boston Dynamics, the majority of (research) robots that resemble humans or animals came from Japan.
But paradoxically, I think the reason is that they view tech less like a tool - all things have spirits in Japanese Shintoism, right? Perhaps therein lies the paradox: maybe due to this, the idea of AI becoming self-aware is less scary - objects are already treated as having a "spirit" anyway!
I think the story of Roujin Z contains a very interesting example of this[0]. An old man becomes (unwilling) part of an experiment for robots designed to take care of the elderly. The robot then becomes a rampaging AI, which turns out to be because it's secretly a government experiment for re-arming Japan. Due to uploading the voice of the wife deceased of the old man, the AI also takes on the persona and becomes a caring (but still rampaging) AI.
Let me emphasize: only the old lady's voice is uploaded - which important because it implies anthropomorphized magical thinking: like Zhang's dragon coming to life when he gave it eyes[1], giving the AI the voice of a sweet old lady magically made it a sweet old lady.
So if we go back to generalising, I think the difference is that Westerners fear tech, with stories of AI turning against us, while Japanese people consider it a potential new friend. Making them much more enthusiastic about new technology.
They were NSA employees and IBM engineers. There are no blanket prohibitions on travel for NSA folks, although I believe there are more or less 'don't go there' advisories about specific places. I also imagine there are folks there who know things that put them in special travel categories.
I don't think the IBMers had any special conditions imposed on them.
This is one of those modalities (gait, keystrokes, tone of voice) that gets rediscovered every few years, shown to be waaay worse than e.g. iris recognition and forgotten until somebody else picks it up again.
Back when I was doing my thesis(liveness detection based on pupil dilation) there existed a system that could scan multiple irises from a distance of up to 6 metres. I assume these kind of systems have only gotten better since then.