Actually, facial recognition algorythms are already being tested in some shops and in stadiums. I believe those have a small fraction of the processing power that a government could have if they meant serious business.
And nobody said you'd want to watch 100% of the territory, or remote areas, or your toilets. If you have cameras in airports and train stations for a start it's enough. So why do you always reply about the lack of ressources to instal cameras in remote mountain toilet cabins instead of replying to what people actually say?
Nic
Smile
-
Pudfark
Re: Smile
Well, again....It's like this Nic?
Limited to what you say?
It's still not "reasonably" doable.
The why of that is still the same as mentioned before.
The time it will take?
Is much to long.
I don't know why you think?
That an "automated" system is gonna apprehend people.
Reference Soapy's astute remark.
Limited to what you say?
It's still not "reasonably" doable.
The why of that is still the same as mentioned before.
The time it will take?
Is much to long.
I don't know why you think?
That an "automated" system is gonna apprehend people.
Reference Soapy's astute remark.
Re: Smile
I guess this IS the right thread to the paranoid topic, not the other. Didnt read this one earlier, my bad.
Anyway.
This face recognition system gives me the creeps, because it would enable this Orwel's 1984 - dystopia. Or Soviet Union/DDR 2.0 . In both visions citizens are monitored and spied all the time. Say something that goverment thinks suspicious. Meet people that goverment think suspicious. Do (legal) things too much, and goverment will get suspicious.
With all this automation in place, they don't need the manpower to watch every film. No. They just need to assign clever computers to analyse the data and isolate interesting patterns from it. Then the experts can have a closer, intimate look. Too much data you say? Let's make some assumptions.
A camera + device makes an ID number out of a face, a figure of 2^40 should give us enough numbers to ID everyone. That means 40bits = 5 bytes per face. With timestamp and other info added, assume 16 bytes. This ID number is calculated out of facial features, shape of the nose, distance of eyes etc - not a hard thing to do.
Lets assume that there goes on average 1,000 people by the camera per hour. 24,000 times 16 bytes per day - 384,000bytes of raw data. About as much data as in a good porn pic.
Back in 2007 Google processed daily 1,125,899,906,842,624 bytes of data. That divided by 384,000 gives us 2,932,031,007 cameras processed by day. If there's an error in my assumptions, divide that number further. But that should be roughly ten cameras per U.S. citizen.
So the computational capacity isn't the limitation here. Imagination most likely is.
And about sparse population density? My home contry has population density of 16 people / sq. km , U.S. has 33. About vast areas full of nothing? I just spent months on a motorcycle with a tent going around and in Australia. I think I have an idea. Although, I've only visited east and west coasts of US - never spent time at the mid.
Yeah. One little step at a time, freedom gets smaller. I hope I'm wrong on this.
Anyway.
This face recognition system gives me the creeps, because it would enable this Orwel's 1984 - dystopia. Or Soviet Union/DDR 2.0 . In both visions citizens are monitored and spied all the time. Say something that goverment thinks suspicious. Meet people that goverment think suspicious. Do (legal) things too much, and goverment will get suspicious.
With all this automation in place, they don't need the manpower to watch every film. No. They just need to assign clever computers to analyse the data and isolate interesting patterns from it. Then the experts can have a closer, intimate look. Too much data you say? Let's make some assumptions.
A camera + device makes an ID number out of a face, a figure of 2^40 should give us enough numbers to ID everyone. That means 40bits = 5 bytes per face. With timestamp and other info added, assume 16 bytes. This ID number is calculated out of facial features, shape of the nose, distance of eyes etc - not a hard thing to do.
Lets assume that there goes on average 1,000 people by the camera per hour. 24,000 times 16 bytes per day - 384,000bytes of raw data. About as much data as in a good porn pic.
Back in 2007 Google processed daily 1,125,899,906,842,624 bytes of data. That divided by 384,000 gives us 2,932,031,007 cameras processed by day. If there's an error in my assumptions, divide that number further. But that should be roughly ten cameras per U.S. citizen.
So the computational capacity isn't the limitation here. Imagination most likely is.
And about sparse population density? My home contry has population density of 16 people / sq. km , U.S. has 33. About vast areas full of nothing? I just spent months on a motorcycle with a tent going around and in Australia. I think I have an idea. Although, I've only visited east and west coasts of US - never spent time at the mid.
Yeah. One little step at a time, freedom gets smaller. I hope I'm wrong on this.
All I ever wanted, was my own way
Re: Smile
You are not wrong.
Next thing you know the RFID chip is the next cool thing to either get in and ouf of night clubs or to pay without worrying about having your credit card and what not... they won't even need all that camera mumbo jumbo.
Nic
Next thing you know the RFID chip is the next cool thing to either get in and ouf of night clubs or to pay without worrying about having your credit card and what not... they won't even need all that camera mumbo jumbo.
Nic
Re: Smile
http://www.cbsnews.com/2100-215_162-7237750.html
Megacorporations cyberpunk/shadowrun style... or rollerball style if you prefer.
N
Megacorporations cyberpunk/shadowrun style... or rollerball style if you prefer.
N