Is facial recognition set to explode as a security measure

Well, they help the good customers. Respectively:

  • Customers who do lots of business gets better treatment.
  • Customers who don’t steal things get a better experience and lower costs.
  • Customers who legitimately have very specific food requirements get more attentive waiters and customers who aren’t trying to scam free meals get lower costs.
  • Customers who tip well get better tables.
  • Customers who don’t abuse the return policy can get a better return policy because it doesn’t have to account for the abusers.

The last one’s going to happen anyway when they run the customer’s credit. So arguably this could be a bonus for the customer with good credit and the customer with bad credit. They both get to waste less time filling out forms for a credit check.

I too was prepared to point out the silliness of some conspiracy theory that facial recognition systems were all wired with bombs…

My thought was "So, what? You fail the recognition check and a shaped charge in the machine blows your head off? And you think this is standard?"

Sorry to revive this old topic, but it appears to be the most recent thread. The NYT’s “The Daily” podcast had a very interesting show about this yesterday:

If you aren’t into listening to podcasts, they also have a link to the transcript there.

Spoiler alert: If you’ve used social media much, they can figure out who you are very quickly.

In NJ the attorney general has suspended the use of that app by law enforcement until there is a legal review by his office. Before that it was our prosecutors office had already put out that facial recognition software could not be used as probable cause but could be used to develop a suspect that needed to be verified by other means.

I’m not a law-speak-guy but personally I don’t see a legal problem from a law enforcement point of view. It uses pictures and information that the individual is putting out in public. I don’t see how there is an expectation of privacy for pictures and information an individual voluntarily releases in public.

I do understand there are some civil challenges because social media companies want to be paid for using their content even when it is available to the public for free. I’m not sure of the precedent there.

Another recent New York Times article, about the use of facial recognition software in public school security.

It used to be eyes’ iris patterns for solid ID - but which James Bond film had a black hat using a stolen eyeball?

I can’t see mandatory chipping of humans - too insecure. Haven’t stolen high-value dogs been re-chipped?

I don’t know if he originated this idea but an element of John Varley’s novel The Golden Globe has security ports at chokepoints like transport access taking a small flake of skin from a subject’s hand for instant DNA analysis and identification. Spycam facial ID is a hands-off approach and yes, any system can be skirted. Even an un-avoidable ID system can and will be hacked. Anyway, a perfect ID systems provides for total government control. I won’t visit China anytime soon but the US is going that way - so China will come to me. Yikes. Hope I’m dead by then.

How accurate is it really? I had to turn off facial recognition in my Facebook account because it so often mistakenly tagged my son as me.

local cops just pulled the plug on it

Raleigh police abruptly end use of controversial facial recognition tech

Often not that accurate.

The premise of consent crumbles when you use facial recognition software in public schools.

https://www.nytimes.com/2020/02/06/business/facial-recognition-schools.html

~Max

I am even more with voice recognition. My bank recognizes my voice when I phone, even after a stroke affected my diction. Ad this is one of the ten biggest banks in the world, with 20-million account holders.

My assmption is that it can’t identify positively from millions in a database, but if it has two pictures of an identity claimant, it can say That’s close enough.

I’ve worked in computer systems at a large bank system (not in this area, but fellow workers did).

I believe you are preliminarily identified by the phone you are calling from*, then the voice recognition software just compares your voice on the phone to the recorded voice for your account, without having to search the voice records for 20-million account holders. Just verifying that the voice on the phone is a better than x% match to one of a few records is a much simpler (and more accurate) task.

*Most customers only call the bank from a few phone numbers (home landline, cell phone, spouse’s cell phone, etc.). These are stored with your account records. And the bank phone system uses caller id, so they can quickly find your account, as soon as they answer the call.

^^^^ I call using GooglePhone, from overseas, so they never get a repeat originating number. But the rest of your answer makes sense, and is as I suspected.

Or even better (because these systems are horrible at recognizing black people and other people of color), “come into my store and risk getting arrested for something that some other person was accused of sometime between 2 and 20 years ago! You’ll long for the days where we merely followed you around the store like a criminal!”

Last month we had an employee meeting at my work, and they announced that, since the fingerprint system was maxed out(?), they would be going to facial recognition for clocking in and out at each shift.

My face hasn’t been scanned yet.