How advanced is Facial recognition, and 3D printing?

[Seydlitz_k]

Me and my friend were having an argument about Biometrics and their use in security. The core of the argument was that he believed that in 10-15 years we would probably be using facial recognition technologies for money transactions, instead of systems like Chip & PIN.

I disagreed mainly on the basis that the technology would still not be advanced and reliable enough in 10-15 years to be in such widespread use (I know it is already in use in small scale environments, such as some airports, or with some police forces. The UK is also issuing new passports with biometric data in them). Not to mention the fact it would take ages just to create new databases and a new infrastructure capable of supporting a facial recognition system for payment (Things like a reliable terminal machine, a network with enough bandwith, etc.). Therefore, I claimed that it would take at least 20-30 years for something like that to wholly replace current systems. If there was any benefit to doing it at all.

I was also having a similar argument with another friend about 3D printing technology. He claimed that even today, 3D printers can replicate themselves (I did some research. The RepRap can reproduce 60% of itself, but the 40% that it can't produce are the important parts, such as motors and microchips. And it still needs to be assembled), and that in 10-15 years we would all have 3D printers in our homes and we could print out everything we needed at home. Again, I disagreed saying that it's probably take something like 30 years or more to get to the level that we could download and print our own working Nike's.

Whos estimates are more realistic? Am i being to pessimistic about how advanced these technologies are, and how long it takes to develop the infrastructure for something like a facial recognition system for money transactions? I'm asking you guys, since we are a bunch of animators and 3D modelers. It's not really our field.

[Singular Intellect]

Sounds like an overly complex system to use when, if you intent to change the system, something like reading thumb prints would be much simplier and easier to implement.

[Broomstick]

We already have fingerprint recording and reading technology. It's available to the public now. It's also on my government-issued handheld I have from the Census Bureau.

[TheLostVikings]

Me and my friend were having an argument about Biometrics and their use in security. The core of the argument was that he believed that in 10-15 years we would probably be using facial recognition technologies for money transactions, instead of systems like Chip & PIN.

This will most likely never happen (yes i know never is a really long time, but bear with me here) for a few obvious reasons. Like what happens if you are broke and fall down a set of stairs on the way to the bank to get your paycheck, or what if you accidentally have your face "remodeled" in an unfortunate bar brawl and then subsequently don't get recognized when you try to pay for you hospital treatment. And if your facial recognition tech has enough leeway to tackle rather large changes like that (not to mention the constant changes due to aging) it would seem to be vulnerable to "face-spoofing" by plastic surgery.

Facial recognition is decent for say, picking people out of a crowd from a distance, but horrible for any real security. The only current biometrics I have any faith in whatsoever is retinal scans and 3D IR imaging of the veins in the palm, which unlike thumbprints doesn't work if the hand is separated from its owner. (of course the latter means that the carjacking joyriders might kill you after they get to their destination instead of "just" snipping your thumb off, so who knows if that is really such a good feature...)

And they still can't hold a candle to most "Chip & PIN" methods, for one thing the electronic handshakes that validate current credit cards and the like can be constantly updated on both ends to regain security if an exploit is found. I don't really want to have my face "updated" to improve security, if you know what i mean...

[Hawkwings]

I think an important point to bring up here is that multiple systems working together is going to be more secure than a single super-duper ID test. If a security system features a retinal scan, a voice cue, a keypad password, and a card swipe, that's four separate things they have to steal/spoof. And perhaps another layer of security being the order which you do those tests. A lot harder to fool four separate systems than one single security gate.

[Sarevok]

I think there seems to be some bias against face recognition here....

Facial recognition is and never will be perfectly correct every time.

But neither is humans in facial recognition tasks.

As long facial recognition is consistently more accurate then their human counterpart there is no reason why it can not be used more and more in government and security applications.

[Shroom Man 777]

It's odd, but in these shitty district hospitals in the provinces, the employees use frickin fingerprint scanners to clock in for work. Which is really weird, for them to have that kind of stuff while the rest of the hospital's facilities are total shit.

[aimless]

3d printers in homes pumping out everything we need sounds completely ridiculous, because most things we have can't be made that way. You'd have to have some crazy ass 3d printer combo'd with an robotic assembly machine and such a wide multitude of materials and parts...yeah I don't see that happening.

However, if 3d printers become cheap and common, I can see a lot of products being designed specifically so that people can make them in their 3d printers: homogenized materials, design for easy assembly by the buyer, avoiding difficult forms etc. So while you definitely won't have anything close to a replicator, it'll probably be able to make some neat and varied stuff.

[PeZook]

If anything, you will still need to buy raw materials for your MagiTech replicator 3D printer, even if all the other hurdles like nanoscale manufacturing processes for microchips can somehow be elliminated by Quantum.

[Kodiak]

I think that facial recognition is problematic since a person's face can be changed through weight gain, facial hair, or surgery. Also, those of us who are identical twins (1 in 250 births) would have problems since there's somebody out there with the same log-in

I think it more likely to have finger-print scanners or even retinal scanners (ala Minority Report) than to use an expensive facial recognition system.

[Logical Mike]

They are actually moving past facial recognition now. I've received some white pages on access control for the data center that check the top of the hand, apparently that's now where they are going with this as facial recog software is crazy expensive and at best, cumbersome to manage.

In fact, our current model suggest we take new pictures every 6 weeks for reference. Which is fine if you have 40 people, not so much when the number approaches 3000.

[Starglider]

As other posters have noted, facial recognition is not an appropriate choice for financial authorisation no matter how good the software is - there are much better biometric options. However the technology is still progressing fast, is being massively deployed for surveillance applications, and will probably eventually be deployed for personalised advertising and low-priority access control.

3D printing will improve but current techniques are inherently incapable of producing fine tolerance rotating parts, specialised materials (e.g. fuel cell catalysts) and ICs of useful density. Integration of additional robotics and CAM hardware will solve the former issue and also provide self-assembly, but that's a way off. E-beam litho could potentially be minaturised a bit more for small-run and one-off chip production, but I can't see it becoming something an individual hobbyist could afford with anything resembling current techology. Really for the whole Von Neuman experience we need nanoassemblers. However these technologies are complementary with early nanotechnology; the first assemblers will be sharply limited in capability, and will be used for key components only, with more conventional CAM & 3D printing techniques used for the bulk of the assembly.