Author Topic: Why don't surveillance cameras call the cops when a robbery is in progress?  (Read 9628 times)

0 Members and 1 Guest are viewing this topic.

Offline JoeN

  • Frequent Contributor
  • **
  • Posts: 990
  • Country: us
  • We Buy Trannies By The Truckload
I ask this jokingly, but with a real purpose.  When will machine vision become advanced enough that this is a possibility?  Surely it is not that far off now that we have self-driving cars and "IoT", right?  There could be a standard where local police departments allow video feeds over the Internet, all they need to do is supply their IP address (or addresses) to local parties and those parties can provide a push-feed to that IP address (or addresses).  The feed could be full time or part time.  The feed could have a priority.  So there could be an "alarm" feed where the sender pushes it with a warning that a possible burglary or robbery is in progress if the machine vision algorithm determines it is the case or a robbery button is pushed.  The cops can then look at it and make a judgment call.  The security camera, if sufficiently advanced, could zoom it to try to get identifying marks as well as closeups of the face, hands, tattoos, logos on clothes and shoes, etc., and other very high resolution images that could be used to match to apprehended suspects.  Right now, most security footage I see it very poor, and the camera can't call the cops because it is stupid. 

Is there a product here?

« Last Edit: May 03, 2017, 11:39:29 pm by JoeN »
Have You Been Triggered Today?
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 12055
  • Country: us
You can already get systems where an alarm will trigger actual humans to monitor the situation, they will do audio at least and I'd guess video capable systems are available by now, for a monthly fee of course. I think it will be a long time before an automated system can reliably call the police to report an intrusion. It doesn't take many false alarms before the cops get really fed up with you.
 

Online ataradov

  • Super Contributor
  • ***
  • Posts: 6506
  • Country: us
    • Personal site
That's what security companies are for. You are assuming that PDs have staff to watch all those videos, they don't. They barely have enough resources to respond to things that are getting reported, they don't need more work.

And even with real security companies - 3-4 false alarms in a row in a short amount of time, and your next alarm gets lower priority than kitty stuck on a tree.
Alex
 
The following users thanked this post: Kilrah

Offline JoeN

  • Frequent Contributor
  • **
  • Posts: 990
  • Country: us
  • We Buy Trannies By The Truckload
You can already get systems where an alarm will trigger actual humans to monitor the situation, they will do audio at least and I'd guess video capable systems are available by now, for a monthly fee of course. I think it will be a long time before an automated system can reliably call the police to report an intrusion. It doesn't take many false alarms before the cops get really fed up with you.

I think a working system would be more like "something odd is going on here, please take a look at it" rather than "alarm" with automatic dispatch.  There are always people manning 911, having them look over some video feeds doesn't seem to me to be asking too much of them.  With Internet and WiFi everywhere now, I think a standard could be made for this sort of video feed really easily that would allow a sender (convenience store, someone on vacation, etc.) and receiver (local police) to set up communications with maybe as few parameters as the receiver supplying an IP address.  I would think the cops would love this, maybe for the wrong reasons, but because it gives them potentially a lot of free cameras around a city to help them track down suspects.

3-4 false alarms in a row in a short amount of time, and your next alarm gets lower priority than kitty stuck on a tree.

I am making maybe the naive assumption here that the algorithms are debugged and generally reliable prior to this being deployed.
« Last Edit: May 04, 2017, 12:20:09 am by JoeN »
Have You Been Triggered Today?
 

Online xrunner

  • Super Contributor
  • ***
  • Posts: 4783
  • Country: us
  • hp>Agilent>Keysight>?
That's what security companies are for. You are assuming that PDs have staff to watch all those videos, they don't. They barely have enough resources to respond to things that are getting reported, they don't need more work.

Yep, they don't have the personal to monitor and respond to video feeds. How many feeds are there going to be hundreds ... thousands? Not going to happen.  ???

I am a Test Equipment Addict (TEA) - by virtue of this forum signature, I have now faced my addiction
 

Offline JoeN

  • Frequent Contributor
  • **
  • Posts: 990
  • Country: us
  • We Buy Trannies By The Truckload
That's what security companies are for. You are assuming that PDs have staff to watch all those videos, they don't. They barely have enough resources to respond to things that are getting reported, they don't need more work.

Yep, they don't have the personal to monitor and respond to video feeds. How many feeds are there going to be hundreds ... thousands? Not going to happen.  ???

Like I said, the point is that the algorithm is smart enough not to pass it on unless someone with the intelligence of a 6 year old or so thought there was a robbery in progress.  So, there are not going to be hundreds or thousands of feeds.  Other than that extreme hyperbole, good point.   :palm:
Have You Been Triggered Today?
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 12055
  • Country: us
I think you'd find that 911 operators are generally pretty heavily loaded, they're not sitting around looking for something to do between calls. They certainly don't have time to monitor security feeds, as someone else said, that's what security system companies are for. They are paid monthly by those who want the service, they can have the necessary staff to monitor feeds from any triggered alarms, that's what they're there for.

There are dozens of break-ins going on across the country at pretty much any given moment, and many, many more false alarms.
 

Online ataradov

  • Super Contributor
  • ***
  • Posts: 6506
  • Country: us
    • Personal site
Like I said, the point is that the algorithm is smart enough not to pass it on
Even regular people in security companies send a lot of questionable reports to the police. I don't see AI being any better.

PS: I listen to a local police radio as a background. The amount of crappy reporting they have to deal with is astounding. There are security companies that fail to articulate where in the building the camera is located.
Alex
 

Offline JoeN

  • Frequent Contributor
  • **
  • Posts: 990
  • Country: us
  • We Buy Trannies By The Truckload
These are not feeds that need to be monitored.  They are feeds that come in when the algorithm, for example, notices extremely unusual situations, for example, masked men with guns in a convenience store.  How hard is this to understand?  It's a huge machine vision algorithm challenge, I understand that, but won't it be doable soon?
Have You Been Triggered Today?
 

Online xrunner

  • Super Contributor
  • ***
  • Posts: 4783
  • Country: us
  • hp>Agilent>Keysight>?
Like I said, the point is that the algorithm ...

Algorithm? What algorithm is going to recognize a robbery? I can understand a vision recognition system that could notice when people are moving in a room, but how does it know it's a robbery? You're talking about AI, the kind we just don't have yet.  :-//

All you really have then is a basic alarm system that notifies {somebody} when the room is occupied because it sees motion. That sort of thing has been around for years - it's called an alarm system.

I am a Test Equipment Addict (TEA) - by virtue of this forum signature, I have now faced my addiction
 
The following users thanked this post: Kilrah

Online ataradov

  • Super Contributor
  • ***
  • Posts: 6506
  • Country: us
    • Personal site
but won't it be doable soon?
Nope. By my estimation more than half of the outside security cameras in downtown San Jose are disabled. Every time something happens police go to the store owners, and cameras are either disabled, or recording to some locked up box with no easy access. Store owners are too cheap to install expensive equipment. They are covered by insurance, even if they install bare minimum, and that's what they do.
Alex
 

Offline Richard Crowley

  • Super Contributor
  • ***
  • Posts: 4310
  • Country: us
  • KE7GKP
  • There are places where systems can call the police. But they require annual permits which can be revoked for excessive false-alarms.
  • Most police departments won't take automatic alarms because they can't handle the volume of false alarms.
      That is why most alarms are connected to monitoring companies who decide what is real enough to call the authorities.
  • There is an alarm company in the US who claim a very high success rate because they use real-time audio monitoring to hear what is actually happening (vs. simply triggering a sensor)
  • Facebook is reported to be hiring 3000 humans to monitor live videos to prevent publicizing crime and violence.
 

Offline stj

  • Super Contributor
  • ***
  • Posts: 2156
  • Country: gb
the camera's arent there to protect you, they are there to watch you because the government fears you.
let a small crowd form infront of a camera and see how fast you get a response!
 

Offline X

  • Regular Contributor
  • *
  • Posts: 184
  • Country: 00
    • This is where you end up when you die...
I can speak from having experience in this field, albeit in Australia. A number of commercial alarm systems exist here which can capture video and send it off to a control location.

A camera's motion detect is unreliable for "sounding an alarm", due to noise from CMOS sensors (particularly noisy in the dark) and moving leaves/wind/etc, and the main purpose of the motion detect on cameras is to avoid recording static scenes. There are cameras which do come with PIRs built into them, but network latency makes them unreliable.

On sites where such security is an issue, the cameras are accompanied with a PIR. The camera has motion detect for the purposes of saving valuable HDD space, while the PIR is there to activate when a someone moves in front of it. The panel can then pull a picture from the camera and send the images to the control room or the client.

I've not seen any (in Oz at least) that calls police directly, they'll often call the monitoring centre first, who will then call the client or contracted security (guard) company to ask if they want the police to be called.
« Last Edit: May 04, 2017, 02:26:11 am by X »
 

Offline Len

  • Frequent Contributor
  • **
  • Posts: 515
  • Country: ca
Algorithms!


 
The following users thanked this post: janoc, tooki, Kilrah

Offline JoeN

  • Frequent Contributor
  • **
  • Posts: 990
  • Country: us
  • We Buy Trannies By The Truckload
Algorithm? What algorithm is going to recognize a robbery?

Hence my initial caveat - "When will machine vision become advanced enough that this is a possibility?"  I know I didn't put that in the headline, it wouldn't fit.

Algorithm? What algorithm is going to drive a car reliably?

I am actually surprised there are so many naysayers on this particular forum.  It's coming.  It may not come as fast as we (or Arthur C. Clarke) expected, but these things are coming.
« Last Edit: May 04, 2017, 05:21:08 am by JoeN »
Have You Been Triggered Today?
 

Offline X

  • Regular Contributor
  • *
  • Posts: 184
  • Country: 00
    • This is where you end up when you die...
I ask this jokingly, but with a real purpose.  When will machine vision become advanced enough that this is a possibility?  Surely it is not that far off now that we have self-driving cars and "IoT", right?
Self driving cars = more accidents.. As for "IoT", well around 80% of all "IoT" devices are junky and useless. Just because something is "IoT" or "smart" or "self-driving" etc, doesn't mean it is useful, and certainly doesn't mean it should be there. The only reason for creating these devices is money and marketing. Nobody in their right mind actually needs them.

There could be a standard where local police departments allow video feeds over the Internet, all they need to do is supply their IP address (or addresses) to local parties and those parties can provide a push-feed to that IP address (or addresses).  The feed could be full time or part time.  The feed could have a priority.  So there could be an "alarm" feed where the sender pushes it with a warning that a possible burglary or robbery is in progress if the machine vision algorithm determines it is the case or a robbery button is pushed.  The cops can then look at it and make a judgment call.
This sort of thing is great on paper, but unrealistic in practice, and the consequences for a false alarm increase significantly.

The security camera, if sufficiently advanced, could zoom it to try to get identifying marks as well as closeups of the face, hands, tattoos, logos on clothes and shoes, etc., and other very high resolution images that could be used to match to apprehended suspects.  Right now, most security footage I see it very poor, and the camera can't call the cops because it is stupid.
Even the most advanced alarms will false-alarm once in a while. I worked on a top-notch installation, high quality cameras and a name-brand alarm system, where the PIRs were being triggered in a regular pattern, which worried the owners as it looked like someone was loitering around waiting for an opportunity to rob the place. Looking through CCTV footage proved difficult due to the noisy picture (night, low illumination) but after enough time, we found that the perimeter PIRs were being triggered in a regular pattern by an energized kangaroo loitering about in the distance (sufficient enough to be difficult to catch on the cameras in most cases). IIRC, some sensitivity adjustments were made, but the roo was ruthless. The issue sort of stopped on its own.
« Last Edit: May 04, 2017, 05:52:24 am by X »
 

Offline Jeroen3

  • Super Contributor
  • ***
  • Posts: 3500
  • Country: nl
  • Embedded Engineer
    • jeroen3.nl
I don't know. Such things have been fantasized about.

But in the real worlds, such intelligent automatic security system would be more expensive than the stores and banks can afford.
Obviously, you can protect your local gas station with the budget of an an air force base, but that a bit out of proportion don't you think?
 

Online Brumby

  • Supporter
  • ****
  • Posts: 10049
  • Country: au
Hence my initial caveat - "When will machine vision become advanced enough that this is a possibility?"

I noticed this from the outset.  The question was never about the present - it was about the future.

So the question is: WHEN?
 

Offline X

  • Regular Contributor
  • *
  • Posts: 184
  • Country: 00
    • This is where you end up when you die...
I don't think we're going to get to "cameras calling the cops" at this stage, and certainly not in the future, but the potential is certainly there. There are high-res cameras (30MP and above) and video analytics packages from companies like Avigilon and iOmniscient, which can either learn what specific behaviours mean, or can be taught to "suspect" activity based on behaviours. These have applications in very large-scale environments at this stage, because each unit costs AU$15000 plus.

I personally have not used anything from iOmniscient, and I have only played around with demonstrations of one of Avigilon's cameras (a 30MP 6fps model), but given enough processing power, and given enough time, this will be very well refined. I don't think the cost is due to the actual manufacturing and materials, but rather the R&D, and in order to sell really new tech, you need to have a high profit margin to pay off the R&D costs.

So the answer to when is when the tech is accessible enough to increase the demand.
« Last Edit: May 04, 2017, 07:21:28 am by X »
 
The following users thanked this post: Kilrah

Offline Kilrah

  • Supporter
  • ****
  • Posts: 1857
  • Country: ch
Algorithm? What algorithm is going to drive a car reliably?
I am actually surprised there are so many naysayers on this particular forum.  It's coming.  It may not come as fast as we (or Arthur C. Clarke) expected, but these things are coming.

Well I guess you should invest into and start working on the development of this then.

IMO it will take more than 10 years until something like this could work semi reliably. Good luck making the difference between a robber breaking in the main door and me coming in in the middle of winter with a thick coat and hat covering my face just as much as a robber would do. Maybe you can differentiate by the robber moving further in the house than me stopping in the hall and drop the heavy clothes, but then the day where I go straight across the whole flat to the kitchen in order to drop whatever heavy stuff I'm carrying you'll be triggering a false alarm.
You can then use my phone's presence to authenticate me, but then I might have forgotten it in the car or have turned wifi/BT off for some reason or they might just not connect becasue we all know the reliability of this. So you need to recognise the car. The whole thing will grow to be highly complicated and way too expensive for anyone to buy into. When things become complicated they become prone to failure too. And when word gets out of how it works the robbers will know they just have to steal your car first to get in without issues.
 

Offline CJay

  • Super Contributor
  • ***
  • Posts: 3547
  • Country: gb
I'm not convinced we will ever see fully automatic systems that can more reliably detect a robbery in progress faster than a cashier, seems to me that if there's a human factor involved then that's going to be far more reliable unless there's some element of extreme violence involved to prevent them pushing an alarm button.

I can see the potential need in a fully automated environment but there are better and simpler ways to detect robbery of an ATM or vending machine for instance.
 

Offline NivagSwerdna

  • Super Contributor
  • ***
  • Posts: 2052
  • Country: gb
If the system can reliably identify intruders then logically it should just activate the machine gun. (castle law)

No need for the cops

 ;)
 

Offline amyk

  • Super Contributor
  • ***
  • Posts: 6816
Hence my initial caveat - "When will machine vision become advanced enough that this is a possibility?"

I noticed this from the outset.  The question was never about the present - it was about the future.

So the question is: WHEN?
I hope never. Pervasive surveillance is creepy enough as it is already. :-\
 

Online xrunner

  • Super Contributor
  • ***
  • Posts: 4783
  • Country: us
  • hp>Agilent>Keysight>?
Hence my initial caveat - "When will machine vision become advanced enough that this is a possibility?"  I know I didn't put that in the headline, it wouldn't fit.



I am actually surprised there are so many naysayers on this particular forum.  It's coming.  It may not come as fast as we (or Arthur C. Clarke) expected, but these things are coming.

Your thread title -

Why don't surveillance cameras call the cops when a robbery is in progress?

I answered it - because the technology doesn't exist for a machine to analyze a moving scene to tell if a human being is committing a robbery. It simply does not exist NOW.

In the future - OF COURSE!

I believe it will exist someday and I hope it will exist someday. But right now it doesn't. Come on JoeN!

Quote
Algorithm? What algorithm is going to drive a car reliably?

I don't know. I don't write self-driving car software.

I don't think that's the exact question you really wanted to ask me though ... what you need to do is think about the questions you are asking a little bit longer and write the questions a little bit clearer.  :-//
I am a Test Equipment Addict (TEA) - by virtue of this forum signature, I have now faced my addiction
 
The following users thanked this post: jpc

Online NANDBlog

  • Super Contributor
  • ***
  • Posts: 4648
  • Country: nl
  • Current job: ATEX certified product design
OK, so imagine this. This is in USA. Robbery going on, camera warns the police department, they go on the scene, robber pulls out a gun, barricades themselves. Instead of a robbery, and some few thousand dollar damage (oh no, they took the jewelry, that is really really important because it is from gold) , you put police officers at risk, probably damage your house, maybe turn it into a hostage situation, maybe car chase, dead people. Nobody likes that. Besides, what's the point? Insurance will pay anyway.
There was this guy prosecuted just now. Someone was stealing his cucumber from his garden, so he wired up his fence to the 230V, thief died. Justice served, right?

But I know, USA is different. Every life is precious, except if it steps on my property!
 

Offline Halcyon

  • Global Moderator
  • *****
  • Posts: 3967
  • Country: au
I've not seen any (in Oz at least) that calls police directly, they'll often call the monitoring centre first, who will then call the client or contracted security (guard) company to ask if they want the police to be called.

Correct. Camera's don't notify Police. Not only would it be unreliable, it wouldn't be able to distinguish between a person who isn't meant to be on-site and a keyholder etc...
 

Offline madires

  • Super Contributor
  • ***
  • Posts: 5260
  • Country: de
  • A qualified hobbyist ;)
Yes, that feature will added just after the manufacturers managed to write secure firmwares >:D At the moment your security cam is a proncam for some script kiddies and cyber crooks. And AI is over-hyped. Put a sticker on the traffic sign and the AI fails. The last DefCon or BlackHat had a talk about fooling driver assistance systems. It was surprisingly easy.
 

Offline CJay

  • Super Contributor
  • ***
  • Posts: 3547
  • Country: gb
I've not seen any (in Oz at least) that calls police directly, they'll often call the monitoring centre first, who will then call the client or contracted security (guard) company to ask if they want the police to be called.

Correct. Camera's don't notify Police. Not only would it be unreliable, it wouldn't be able to distinguish between a person who isn't meant to be on-site and a keyholder etc...

Even systems with operators have problems, we used a locker service for delivering parts to mobile techs, one site that hosted a locker has CCTV with operators, a few of our techs have had to explain themselves to the police when the monitoring service has decided they shouldn't be there at 2AM or some other non business hours time.
 

Online BrianHG

  • Super Contributor
  • ***
  • Posts: 3766
  • Country: ca
It's already being worked on, but, in a city wide scale.  See insane camera #2 in this video (at 1min,54sec):
https://youtu.be/cm0KJC3b3UQ?t=114
__________
BrianHG.
 

Offline Bicurico

  • Frequent Contributor
  • **
  • Posts: 991
  • Country: pt
    • VMA's Satellite Blog
Cameras are used to DOCUMENT a crime. After the crime, the footage can be used to help figure out what happend and who did it.

Most of the times however, the resolution, angle, light, etc. is not good enough.

To trigger the presence of a human in unusual hours and hence assuming a robbery is going on, it is simpler to use IR technology. Otherwise existing cheap motorised 360ยบ IP cameras from China will detect significant changes in the video and trigger an alarm with e-mail and attached picture. Just configure it with the local police e-mail and you are set.

The question is: if there is no identified person CALLING the police, these messages will be ignored. You might in fact get a fine for operating such a device without permit!

Regards,
Vitor

Offline Kilrah

  • Supporter
  • ****
  • Posts: 1857
  • Country: ch
Currently yes, but the whole subject of this topic is to imagine that through future development a camera could be able to recognise people, and differentiate "normal" behavior from "suspect" behavior.
 

Offline ebclr

  • Super Contributor
  • ***
  • Posts: 2105
  • Country: 00
suspect" behavior" is a very relative concept

"Those who give up their liberty for more security neither deserve liberty nor security."
 

Jackson

 
The following users thanked this post: tooki, stj

Offline Red Squirrel

  • Super Contributor
  • ***
  • Posts: 2462
  • Country: ca
Idealy calling cops should normally be up to a human.  That's why NOCs exist, you view alarms and make a proper judgement call. 

Unfortunately there are lot of cases where it does not work that way, especially with copyright.  They automate the hell out of that and anything that is even close to copyrighted material gets your stuff censored off the internet.  Imagine the physical version of that.  Get jumped by cops since you walked by a camera with a cigarette and the system thought it was a gun.
 

Offline Bicurico

  • Frequent Contributor
  • **
  • Posts: 991
  • Country: pt
    • VMA's Satellite Blog
If you want to automate suspect behaviour, why not automate prediction of crime?

Using databases and profiling you can ANTECIPATE a crime before it happens and effectively stop the criminal. Don't recall the name of the movie with Tom Cruise that explores this theme.

The question is: do we want to live in such a world?
Do we want to have constant monitoring?
How effective is monitoring to prevent crime? To solve crimes after they happend?

Not an easy subject!

I think it would be easier and cheaper to HELP population to study and educate themselfes, leading them to a prosperous life.

Instead goverments are lowering standards of education, opinion, wages, culture. This ultimately causes crimes to rise.

Wanting to AUTOMATE crime prevention/investigation is ridiculous in this context.

That apart: I say that today's technology would be well suited for automated crime reporting based on behaviour patterns. No big deal in that.

Remember that airports and stadiums have automated person identification systems implemented in the many cameras they have! If a wanted person is filmed, he will be automatically identified and reported!

The future is already here. For good and bad.

Regards,
Vitor

Offline CJay

  • Super Contributor
  • ***
  • Posts: 3547
  • Country: gb
If you want to automate suspect behaviour, why not automate prediction of crime?

Using databases and profiling you can ANTECIPATE a crime before it happens and effectively stop the criminal. Don't recall the name of the movie with Tom Cruise that explores this theme.

No. Absolutely not, that's way too far down the thought crime route.
 
The following users thanked this post: tooki

Offline james_s

  • Super Contributor
  • ***
  • Posts: 12055
  • Country: us
I think it is perhaps reasonable to anticipate crime and pay extra attention to locations and suspects, but I certainly would never advocate taking action against someone who has not yet actually committed a crime.
 

Offline JoeN

  • Frequent Contributor
  • **
  • Posts: 990
  • Country: us
  • We Buy Trannies By The Truckload
I believe it will exist someday and I hope it will exist someday. But right now it doesn't. Come on JoeN!

Like I said, I got to the real question quickly enough "When will machine vision become advanced enough that this is a possibility?"  Sorry that my title was intentionally meant to be catchy.  It caught you as intended, unfortunately it didn't convince you to read the actual text of the message.   |O
Have You Been Triggered Today?
 

Offline JoeN

  • Frequent Contributor
  • **
  • Posts: 990
  • Country: us
  • We Buy Trannies By The Truckload
OK, so imagine this. This is in USA. Robbery going on, camera warns the police department, they go on the scene, robber pulls out a gun, barricades themselves. Instead of a robbery, and some few thousand dollar damage (oh no, they took the jewelry, that is really really important because it is from gold) , you put police officers at risk, probably damage your house, maybe turn it into a hostage situation, maybe car chase, dead people. Nobody likes that. Besides, what's the point? Insurance will pay anyway.
There was this guy prosecuted just now. Someone was stealing his cucumber from his garden, so he wired up his fence to the 230V, thief died. Justice served, right?

But I know, USA is different. Every life is precious, except if it steps on my property!

"Imagine this."  Imagine that.  Imagine Dragons!  You are not even discussing the technology now, you are discussing police strategy and procedure.  And, in fact, if police are notified of a robbery, most of the time they wait for the perp outside for this very reason.  Your great idea here is to not notify police of a robbery in progress because they might make a mistake or choose a boneheaded strategy.  In general, if someone knows that a robbery is occurring, they notify the police in every civilized country.  Doing that with this sort of technology is not dissimilar to doing it with a hidden button or a citizen spotting a robbery and calling it in.
« Last Edit: May 05, 2017, 12:38:13 am by JoeN »
Have You Been Triggered Today?
 

Online xrunner

  • Super Contributor
  • ***
  • Posts: 4783
  • Country: us
  • hp>Agilent>Keysight>?
Like I said, I got to the real question quickly enough "When will machine vision become advanced enough that this is a possibility?"  Sorry that my title was intentionally meant to be catchy.  It caught you as intended, unfortunately it didn't convince you to read the actual text of the message.   |O

Alright I'll try again.

So you are asking this -

When will machine vision become advanced enough that this is a possibility? 

My answer is - I do not know.

I am a Test Equipment Addict (TEA) - by virtue of this forum signature, I have now faced my addiction
 

Offline JoeN

  • Frequent Contributor
  • **
  • Posts: 990
  • Country: us
  • We Buy Trannies By The Truckload
My answer is - I do not know.

You and me both!   |O
Have You Been Triggered Today?
 

Offline X

  • Regular Contributor
  • *
  • Posts: 184
  • Country: 00
    • This is where you end up when you die...
"Imagine this."  Imagine that.  Imagine Dragons!  You are not even discussing the technology now, you are discussing police strategy and procedure.  And, in fact, if police are notified of a robbery, most of the time they wait for the perp outside for this very reason.  Your great idea here is to not notify police of a robbery in progress because they might make a mistake or choose a boneheaded strategy.  In general, if someone knows that a robbery is occurring, they notify the police in every civilized country.  Doing that with this sort of technology is not dissimilar to doing it with a hidden button or a citizen spotting a robbery and calling it in.
Discussing police strategy and procedure is 100% relevant to this discussion as you are discussing the idea of cameras calling the police.

Tech only becomes useful if it makes life easier (and even that is a stretch), and I can only see this making life difficult for both citizens and police. In order for the tech to be useful, it has to be more than just "ooh! aah!" rubbish.

Cameras ("machine vision") is already advanced enough (high resolution), and software exists that can identify individuals within a crowd. It's a matter of the algorithms, and whether or not it is worth it.
 

Offline Electro Detective

  • Super Contributor
  • ***
  • Posts: 2713
  • Country: au
Why would cops would want to arrest themselves and associates on a heist, if  'smart' cameras are sending them a live feed? 

I suppose they might want to replay the performance later, and assess options to improve on their speed and technique?    :popcorn:

Hmmm, have I understood this post correctly?    :-//

« Last Edit: May 05, 2017, 06:47:14 am by Electro Detective »
 

Offline ebclr

  • Super Contributor
  • ***
  • Posts: 2105
  • Country: 00
I would recommend read 1984, from George Orwell, before even thinking about this anti-liberty idea.

Cameras IA and cops is all you need for a 100% fascist state, us as engineers, must be against this, not looking to make it. It's a no go way

« Last Edit: May 05, 2017, 03:56:31 am by ebclr »
 
The following users thanked this post: tooki

Offline Halcyon

  • Global Moderator
  • *****
  • Posts: 3967
  • Country: au
Using databases and profiling you can ANTECIPATE a crime before it happens and effectively stop the criminal. Don't recall the name of the movie with Tom Cruise that explores this theme.

Yeh this has been tried, even on a manual level with crime mapping (identifying hot spots/hot times etc...) and at best, it's a stab in the dark. Just because crime has happened in a small defined area in the past, doesn't correlate to increased chances of future crime (of course there are many variables which I won't go into on here).

You may as well get a map, randomly throw a bunch of dots on it and ask Police to patrol those areas, it largely has the same effect.

I remember years ago when I was in the state Police attending calls, I never once came across a crime occurring/about to occur in those so called "hot spot" areas. Not a single one. I did come close one time when a car caught alight due to a faulty starter motor that kept running continuously, however driving a Ford is hardly a crime.
 

Online Brumby

  • Supporter
  • ****
  • Posts: 10049
  • Country: au
If you want to automate suspect behaviour, why not automate prediction of crime?

Using databases and profiling you can ANTECIPATE a crime before it happens and effectively stop the criminal. Don't recall the name of the movie with Tom Cruise that explores this theme.

Minority Report
 
The following users thanked this post: SeanB

Offline jpc

  • Supporter
  • ****
  • Posts: 69
  • Country: gb


Quote from: Bicurico on Yesterday at 15:28:38
If you want to automate suspect behaviour, why not automate prediction of crime?



This is already being done by many police departments in the US using analytical software allied with a whole lot of other information and traditional police work. It's called hot spot policing and has been made possible, or at least easier, by the use of modern IT hardware and software. They use it to predict where crime is more likely then not and then target patrols accordingly and it has proven quite successful in some places and not so much in others.

Obviously this is nothing like the Tom Cruise film Minority Report where they were using psychics to predict that a particular person was possibly going to commit a murder so they would arrest them before they killed.
 

Offline Electro Detective

  • Super Contributor
  • ***
  • Posts: 2713
  • Country: au
Many Youtube videos harp that perhaps 1984 has been around since 1884     :popcorn:      :=\ 

The good news for techies is > don't miss the bus installing and servicing cameras, gotta eat while the conspiracies play out, right?   :-+

Vandals and stream hackers will keep you in work till SHTF in 2084 when paranoid people  :scared:  :scared:  will move and live underground in advanced mancaves to get away from.. well, I don't know   :-//  whatever it is their medication mixed with lemming mentality conjures up by then. 

The trillions of cameras on the surface will be watching other cameras and vice versa, and eventually out of boredom and competitive desire will resort to targetting innocent robot street cleaners merely going about their business, for thought crimes.   

OT: Anyone with Hollywood contacts want to help me flog this script for the next sc-fi blockbuster ?
There's a generous commission fee in it    ;D
 





 

Online NANDBlog

  • Super Contributor
  • ***
  • Posts: 4648
  • Country: nl
  • Current job: ATEX certified product design
OK, so imagine this. This is in USA. Robbery going on, camera warns the police department, they go on the scene, robber pulls out a gun, barricades themselves. Instead of a robbery, and some few thousand dollar damage (oh no, they took the jewelry, that is really really important because it is from gold) , you put police officers at risk, probably damage your house, maybe turn it into a hostage situation, maybe car chase, dead people. Nobody likes that. Besides, what's the point? Insurance will pay anyway.
There was this guy prosecuted just now. Someone was stealing his cucumber from his garden, so he wired up his fence to the 230V, thief died. Justice served, right?

But I know, USA is different. Every life is precious, except if it steps on my property!

"Imagine this."  Imagine that.  Imagine Dragons!  You are not even discussing the technology now, you are discussing police strategy and procedure.  And, in fact, if police are notified of a robbery, most of the time they wait for the perp outside for this very reason.  Your great idea here is to not notify police of a robbery in progress because they might make a mistake or choose a boneheaded strategy.  In general, if someone knows that a robbery is occurring, they notify the police in every civilized country.  Doing that with this sort of technology is not dissimilar to doing it with a hidden button or a citizen spotting a robbery and calling it in.
Humans have the ability to decide in moral situations better at the moment than robots. I would let the robbers leave my home before I call the police. What they are taking is just stuff. Sure it is my stuff, but whatever they are taking can be replaced. If they are robbing a bank, sure rob it. I lie on the floor and let the situation go away. Sure in america people think that at the fuel station you need a shotgun under the counter and legally kill everyone who wants to take that 20 bucks in the counter. The rest of the world thinks otherwise.
We are discussing a system which will automatically report a crime in progress. That is one tiny step away from going full Judge Dredd. I suppose the system should also allow the police acess to the cameras. But you know, what? The security system in your house should also report other crimes. For example someone underage smoking marijuana. Live feed to police, so they can arrest him, right away. Or any of the sexual activities that are considered "crimes". I mean, it is just a machine looking, so not spying. You know, if you have a sex toy in Nevada, that is a crime. We should make cameras report ALL crimes, and place it everywhere. And make AIs to analyse or every move, and guide us.
You have nothing to fear. If you are a law obeying citizen.
 
The following users thanked this post: CJay

Offline CJay

  • Super Contributor
  • ***
  • Posts: 3547
  • Country: gb
A utopian attitude perhaps but I hope that one day the world will be fair enough that there isn't a need for robbery and thus make any need for such technology obsolete.

Of course it'll never happen and crimes such as murder, rape, assault etc. will still happen even if there is no need or desire to steal anymore but I can hope can't I?
 
The following users thanked this post: Kilrah

Offline X

  • Regular Contributor
  • *
  • Posts: 184
  • Country: 00
    • This is where you end up when you die...
If the system can reliably identify intruders then logically it should just activate the machine gun. (castle law)

No need for the cops

 ;)
I would never trust a computer to reliably identify intruders and kill them, but I'm happy for the computer to smoke them out of the building:


This has been installed in a few buildings in Australia.
 

Offline CJay

  • Super Contributor
  • ***
  • Posts: 3547
  • Country: gb
Yeah, smokecloak is used here too, used to buy from a PC component supplier who had it, apparently it triggered one night when one of the directors brought some Friday night conquest back to the office with the intention of showing them his 'business acumen' on the boardroom table.
 

Offline Electro Detective

  • Super Contributor
  • ***
  • Posts: 2713
  • Country: au
More evidence how cheap some  'directors' are   :--

Good news the prick got smoked out, the lady can find better action elsewhere  :-+
 

Offline Kilrah

  • Supporter
  • ****
  • Posts: 1857
  • Country: ch
Smoke screens are one thing that's very efficient but for some reason already crosses the acceptance threshold of most... had a friend who tried to distribute some, but regardless of all the demos he did nearly everyone dismissed it. Too expensive, too "complicated", "meh i don't care, there's insurance"...
 

Offline CJay

  • Super Contributor
  • ***
  • Posts: 3547
  • Country: gb
They're very effective and can be mandated by insurance companies, especially if you've had high value prior claims but I believe some leave a residue and are a PITA to clean up after.

 

Online janoc

  • Super Contributor
  • ***
  • Posts: 3108
  • Country: fr
If you want to automate suspect behaviour, why not automate prediction of crime?

Using databases and profiling you can ANTECIPATE a crime before it happens and effectively stop the criminal. Don't recall the name of the movie with Tom Cruise that explores this theme.

No. Absolutely not, that's way too far down the thought crime route.

And, in fact, this is exactly what is being done. Crime prediction software that tells where to send more patrols based on the past records of arrests and crime is in routine use already.

https://www.predpol.com/

Now the issue with it is that it is only as good as its data - so if you have a racist, corrupt cops overpolicing a black/poor/immigrant neighborhood and turning blind eye to crime elsewhere, guess what happens when that data gets fed into the software - "Quarter XY has 75% more arrests than everyone else, so we need to send much more police there!", perpetuating (or even worsening) the situation. This effect of "washing" the institutional biases through the statistics and "AI" software has been also thoroughly documented.

Quote from: NANDBlog
OK, so imagine this. This is in USA. Robbery going on, camera warns the police department, they go on the scene, robber pulls out a gun, barricades themselves. Instead of a robbery, and some few thousand dollar ...

That's BS - when someone picks a phone up and calls the cops "manually" about a robbery (or whatever crime) being in progress, you think they won't show up? What is the difference?

If what you are saying was true, why would police ever respond to something like a terrorist attack or mass shooting? The robber will most likely start to run away when seeing cops, most aren't suicidal and the chances of winning a standoff with the police are pretty much zero. OTOH a terrorist or a nut with a gun are pretty much guaranteed to shoot - and yet the police responds, even though the chances of getting killed are much higher.


To @JoeN:

Despite the enormous hype in media about the advance of AI, do keep in mind that all that is going on right now is one field of AI (machine learning) being used as a hammer on every possible problem.

Even if we had some sort of neural network or some other magical algorithm that would be perfectly able to identify when a robbery is going on as opposed to people talking or having an argument (for ex.), it still wouldn't do anything about the fact that surveillance camera images generally suck badly, even humans have often problem to interpret what is going on in them. And that isn't really changing, especially when it comes to night time scenes.

Humans also tend to err on the side of caution and will use a common sense and their experience when judging when a situation warrants to call the cops. It is not perfect, but it acts as a huge filter for BS results. No machine learning algorithm has anything like that - all they do is using more or less sophisticated math and various heuristics to match the input against what they have seen before.

There isn't anything new there that hasn't been done before, all that is happening is that the availability of huge datasets and computational resources enables to build systems that were not possible before. But there hasn't really been any major breakthrough in understanding the scenes the algorithms "see" or that the algorithms gained some fundamentally new ability they didn't have before.
 
The following users thanked this post: jpc

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 15390
  • Country: za
Smokescreen not used much here, there is a growing use of a pepper spray system, which is very effective. No burglary possible if you cannot breathe, and cannot see due to watering eyes.  plus it makes for good crimewatch footage as well, which probably is the best advertising the sellers have, of places being broken into, but the thieves not getting loot, and being arrested at the local ER with severe inhalation issues.

Most robberies the cheapest thing stolen is the actual goods, the biggest cost is replacing broken glass, smashed doors and such, and then adding yet more razor wire, more electric fencing and stronger security bars. Only ones making out like bandits are those suppliers of razor wire ( I buy local, they offer a full service supply of all you need, from the poles to the wire and even will install at a price), the glaziers and of course the police with the " here is the OB number for your crime, do you want us to investigate it, please contact your insurance, I feel the need for some ' cooldrink' coming".
 

Offline Red Squirrel

  • Super Contributor
  • ***
  • Posts: 2462
  • Country: ca
If the system can reliably identify intruders then logically it should just activate the machine gun. (castle law)

No need for the cops

 ;)
I would never trust a computer to reliably identify intruders and kill them, but I'm happy for the computer to smoke them out of the building:


This has been installed in a few buildings in Australia.

That's a neat idea.

Replace with CO2 and that robber is never going to try robbing a bank again.  :P
 

Online vodka

  • Frequent Contributor
  • **
  • Posts: 527
  • Country: es
Smokescreen not used much here, there is a growing use of a pepper spray system, which is very effective. No burglary possible if you cannot breathe, and cannot see due to watering eyes.  plus it makes for good crimewatch footage as well, which probably is the best advertising the sellers have, of places being broken into, but the thieves not getting loot, and being arrested at the local ER with severe inhalation issues.

Most robberies the cheapest thing stolen is the actual goods, the biggest cost is replacing broken glass, smashed doors and such, and then adding yet more razor wire, more electric fencing and stronger security bars. Only ones making out like bandits are those suppliers of razor wire ( I buy local, they offer a full service supply of all you need, from the poles to the wire and even will install at a price), the glaziers and of course the police with the " here is the OB number for your crime, do you want us to investigate it, please contact your insurance, I feel the need for some ' cooldrink' coming".

If you believe that with  an electric fence and razor wire  you will stop the burglars. You have it crude crude. They could rob a car and clash versus electric fence or could cut the fence with portable  grinder  or shears.
Here , the copper burglars steal the  wires of catenaries with portable grinder wrapped to long  wooden stick  .
 

Offline Jeroen3

  • Super Contributor
  • ***
  • Posts: 3500
  • Country: nl
  • Embedded Engineer
    • jeroen3.nl
Copper thieves rarely steal railway overhead line. They don't dare to touch high voltage DC yet.
However, they do steal the control cables weekly. A lot of effort is being put in replacing those long cables for fiber.

But ripping all gas pipes from office buildings at night is nothing new.
 

Online NANDBlog

  • Super Contributor
  • ***
  • Posts: 4648
  • Country: nl
  • Current job: ATEX certified product design
Copper thieves rarely steal railway overhead line. They don't dare to touch high voltage DC yet.
However, they do steal the control cables weekly. A lot of effort is being put in replacing those long cables for fiber.

But ripping all gas pipes from office buildings at night is nothing new.
You would think. Statistics from Hungary. There was a theft in Budapest in broad daylight few years ago. They stole about 100m overhead cable on a tram line which transports half a million people a day, and has trams every 60 seconds or so. Statistics said, they stole 13.5 km of that wiring per year. Train company said, 460 times a year did they find out that cabling was stolen. As I understand, they tried stealing from the 750KV wiring at some point.
Absolutely ridiculous. They changed the law a few years ago, so smelteries cannot accept wiring anymore, only if you have some papers with it. That reduced cable theft by some 75%, continues to decline. It is very easy to fight this.
 

Offline X

  • Regular Contributor
  • *
  • Posts: 184
  • Country: 00
    • This is where you end up when you die...
That's a neat idea.

Replace with CO2 and that robber is never going to try robbing a bank again.  :P
Hopefully there won't be any false alarms. But why kill the robber immediately? Just use a highly irritating gas (eg. fumes from cheap Chinese glue) to give him cancer as he is being smoked out.  >:D

Smoke screens are one thing that's very efficient but for some reason already crosses the acceptance threshold of most... had a friend who tried to distribute some, but regardless of all the demos he did nearly everyone dismissed it. Too expensive, too "complicated", "meh i don't care, there's insurance"...
The standard attitude towards security in general:
ACT I: THE FIRST MISTAKE
OWNER: "Our own abode of trade doth be well protected, I say."
ACT II: A CALL TO ELECTRIC ARMS
OWNER returns to his business, to find a cracked window and damaged/missing goods.
Enter SECURITY FIRM.

OWNER: "I say, a delinquent hath engaged in larceny. My dear security firm, what price must I pay, for cameras, robots, lasers, and a scarecrow that maketh mine enemies afeard?"
(SECURITY FIRM and OWNER walk around all over site)
SECURITY FIRM: "Alas, a hefty price,my liege. Seven hundred shekels of silver, and another fifty."
OWNER: "What?! Why such preposterous pricing?!"
SECURITY FIRM: "Your majesty, thou can get away with the cameras. Just a hundred shekels."
OWNER: "Nay! I hath better not. I shall instill fear into the hearts of delinquents by hanging manly artifacts on mine windows."
SECURITY FIRM: "Your Excellency hast been advised."
ACT III: HMMMPPH
OWNER: "Hmmmpph!"
SECURITY FIRM: "Hmmmpph!"
EVERYONE ELSE: "Hmmmpph!"
OWNER'S DOG: "Hmmmpph!"
EXEUNT

They're very effective and can be mandated by insurance companies, especially if you've had high value prior claims but I believe some leave a residue and are a PITA to clean up after.
Never heard of insurance companies mandating this, but I think some provide bonuses such as lowered premiums for properties with specific security measures.

And, in fact, this is exactly what is being done. Crime prediction software that tells where to send more patrols based on the past records of arrests and crime is in routine use already.

https://www.predpol.com/

Now the issue with it is that it is only as good as its data - so if you have a racist, corrupt cops overpolicing a black/poor/immigrant neighborhood and turning blind eye to crime elsewhere, guess what happens when that data gets fed into the software - "Quarter XY has 75% more arrests than everyone else, so we need to send much more police there!", perpetuating (or even worsening) the situation. This effect of "washing" the institutional biases through the statistics and "AI" software has been also thoroughly documented.
I see no issue with an AI being programmed to focus its monitoring efforts on areas which are statistically shown to require a higher police presence. Police do this to save resources anyway, so all they'll be doing is feeding it into the AI. In practice the AI doesn't do the policing, the police have to expend resources, so they want to achieve this in the most efficient way possible. And if more police are sent to "black/poor/immigrant neighbourhoods" because these areas are shown to have a significantly higher crime rate per head, then that's how it should be. The AI predicts crime based on a variety of parameters, and statistical and known history will naturally influence this prediction.

The robber will most likely start to run away when seeing cops, most aren't suicidal and the chances of winning a standoff with the police are pretty much zero. OTOH a terrorist or a nut with a gun are pretty much guaranteed to shoot - and yet the police responds, even though the chances of getting killed are much higher.
Some shoplifters will even do this.

Even if we had some sort of neural network or some other magical algorithm that would be perfectly able to identify when a robbery is going on as opposed to people talking or having an argument (for ex.), it still wouldn't do anything about the fact that surveillance camera images generally suck badly, even humans have often problem to interpret what is going on in them. And that isn't really changing, especially when it comes to night time scenes.
You should look at footage from modern-day IP cameras. Even the no-name Chinese brands out there have good image quality compared to the "analog" equivalents, and the cost of IP-based CCTV installations have recently dropped to the point where they are accessible even to those on a tight budget.

AI already exists that can identify patterns of certain behaviours, but this are generally used with surveillance cameras in the order of 20 megapixels or above. The frame rate of such cameras is not the best (presently around 6fps), but this is going to get better with time, and surveillance applications usually don't need the same frame rate as a Hollywood blockbuster anyway.

Humans also tend to err on the side of caution and will use a common sense and their experience when judging when a situation warrants to call the cops. It is not perfect, but it acts as a huge filter for BS results. No machine learning algorithm has anything like that - all they do is using more or less sophisticated math and various heuristics to match the input against what they have seen before.

There isn't anything new there that hasn't been done before, all that is happening is that the availability of huge datasets and computational resources enables to build systems that were not possible before. But there hasn't really been any major breakthrough in understanding the scenes the algorithms "see" or that the algorithms gained some fundamentally new ability they didn't have before.
Calling the police over suspected armed robberies may not even be the best issue to tackle using AI, which seems better suited to identifying potentially malicious series of actions occurring over a period of time. One issue particularly in large organisations, is staff pilfering (staff stealing products or finances, fraud caused by staff, etc). In many CCTV installs, there is often at least one camera viewing the till, and staff are denied administrative access to the CCTV system (though they are granted enough to back up footage) for this exact reason. I would be very interested to see a video analytics AI algorithm developed to identify suspicious behaviour, since this kind of thing is a lot more subtle and difficult to notice than a robbery.

Smokescreen not used much here, there is a growing use of a pepper spray system, which is very effective. No burglary possible if you cannot breathe, and cannot see due to watering eyes.  plus it makes for good crimewatch footage as well, which probably is the best advertising the sellers have, of places being broken into, but the thieves not getting loot, and being arrested at the local ER with severe inhalation issues.

Most robberies the cheapest thing stolen is the actual goods, the biggest cost is replacing broken glass, smashed doors and such, and then adding yet more razor wire, more electric fencing and stronger security bars. Only ones making out like bandits are those suppliers of razor wire ( I buy local, they offer a full service supply of all you need, from the poles to the wire and even will install at a price), the glaziers and of course the police with the " here is the OB number for your crime, do you want us to investigate it, please contact your insurance, I feel the need for some ' cooldrink' coming".
It is important to understand that the primary purpose of a security defence is not to cause harm to miscreants, but to protect something against them. In the case of burglars, try to achieve this by disorienting them and wasting their time. Burglars don't like (and often cannot afford) to have their time wasted, because this increases the chance that they will be caught by either civilians or police. The more you waste the burglar's time, the better the outcome for you.
Of course, harming burglars in the process wastes more time by forcing them to tend to their injuries, and can waste even more of their time when they are immobilised as a result, however such harm should not be the primary motive/objective behind a security system or defence. The primary objective of a security system must always be to protect something.

Unfortunately (at least in Australia) I can see a future where even remotely advanced security systems are outlawed. With idiotic, harmful, unrealistic and over-the top anti-discrimination laws, health and safety regulations, gun controls, and limits on how people can defend themselves and their property, the insanely evil and the incurably stupid are well protected and compensated, while the real victims get screwed over badly. It is likely that even more laws will be introduced to protect stupidity, and security systems of the future will just file a record in an event log and do nothing else. Things aren't looking too good for the security industry here, yet alone a future of cameras calling police, and robbery-identifying AI.

We have most of the components to implement the tech, but good luck doing that without a lawsuit from an armed robber who suddenly turns up to court in a business suit and has a lifetime supply of crocodile tears. There are many situations where this sort of thing has happened, and owners have been annihilated in court as a result.
« Last Edit: May 10, 2017, 04:59:36 pm by X »
 

Offline CJay

  • Super Contributor
  • ***
  • Posts: 3547
  • Country: gb
give him cancer as he is being smoked out.  >:D

I hope you or anybody else in your life ever has to deal with cancer. I've not been so fortunate and you're a callous dickhead.
 
The following users thanked this post: SeanB

Online janoc

  • Super Contributor
  • ***
  • Posts: 3108
  • Country: fr
I see no issue with an AI being programmed to focus its monitoring efforts on areas which are statistically shown to require a higher police presence. Police do this to save resources anyway, so all they'll be doing is feeding it into the AI. In practice the AI doesn't do the policing, the police have to expend resources, so they want to achieve this in the most efficient way possible. And if more police are sent to "black/poor/immigrant neighbourhoods" because these areas are shown to have a significantly higher crime rate per head, then that's how it should be. The AI predicts crime based on a variety of parameters, and statistical and known history will naturally influence this prediction.

Are the "black/poor/immigrant neighbourhoods" areas with significantly higher crime rates such because those people are really more prone to be criminals or because those areas are overpoliced, where you can get arrested even for trivial stuff that is ordinarily let slide elsewhere? Especially if you take into account the "broken windows" policing doctrine (= harshly punishing everything, no matter how tiny and insignificant, with the idea that it will discourage more serious crime) that has been widely discredited but is still widely applied.

Now take such data and feed them into a neural network or some other machine learning system. The problem is that if you train that AI with a history of crime that is based on pre-existing biases (immigrant/poor/black/muslim/whatever ...), you will actually only aggravate the situation. Basically a machine-reinforced confirmation bias.

http://www.sciencemag.org/news/2017/04/even-artificial-intelligence-can-acquire-biases-against-race-and-gender
https://www.propublica.org/article/machine-bias-risk-assessments-in-criminal-sentencing
https://www.nytimes.com/2016/06/26/opinion/sunday/artificial-intelligences-white-guy-problem.html?_r=0

It actually gets worse - there are similar software black boxes that will evaluate your likelihood to re-offend and suggest sentence to the judges. Guess who gets hit with much tougher sentences for the same crimes.

There is a good book on why these black box models where we don't know how the system actually arrives to the result a huge problem:
https://weaponsofmathdestructionbook.com/


AI already exists that can identify patterns of certain behaviours, but this are generally used with surveillance cameras in the order of 20 megapixels or above. The frame rate of such cameras is not the best (presently around 6fps), but this is going to get better with time, and surveillance applications usually don't need the same frame rate as a Hollywood blockbuster anyway.

These behavior-based approaches have been widely discredited as an ineffective woodoo, unfortunately they are still being sold by the technology vendors. Such a system would be good to alert the operator that something is going on at best, otherwise it produces way too many false positives.

Calling the police over suspected armed robberies may not even be the best issue to tackle using AI, which seems better suited to identifying potentially malicious series of actions occurring over a period of time. One issue particularly in large organisations, is staff pilfering (staff stealing products or finances, fraud caused by staff, etc). In many CCTV installs, there is often at least one camera viewing the till, and staff are denied administrative access to the CCTV system (though they are granted enough to back up footage) for this exact reason. I would be very interested to see a video analytics AI algorithm developed to identify suspicious behaviour, since this kind of thing is a lot more subtle and difficult to notice than a robbery.

Actually most applications of CCTV are for ex-post investigations - once something happens the footage is actually reviewed.

And re behavior based stuff:

http://www.law.uchicago.edu/files/files/Harcourt%20OpEd%20Behavioral%20Profiling%20Longer%20Version.pdf
https://www.rt.com/usa/tsa-behavior-profiling-program-685/

And that is done by humans - it is very unlikely that some sort of AI would be performing any better than that if even humans have difficulty to predict the problem correctly than a random chance.

[/quote]
 

Offline yada

  • Frequent Contributor
  • **
  • !
  • Posts: 265
  • Country: ca
Like I said, the point is that the algorithm ...

Algorithm? What algorithm is going to recognize a robbery? I can understand a vision recognition system that could notice when people are moving in a room, but how does it know it's a robbery? You're talking about AI, the kind we just don't have yet.  :-//

All you really have then is a basic alarm system that notifies {somebody} when the room is occupied because it sees motion. That sort of thing has been around for years - it's called an alarm system.

You could make it recognize a gun or an out stretched arm with a hand gun in it, or say someone jumping over the counter. Even better most employees wear uniforms so it could look for non uniforms behind the counter. Couple all this with machine learning and I would think its doable. They have cameras on the highways in the US that record and read the licenses plates of *every* single car or truck that comes within so many miles of Washington DC. That's a little diagonal box moving at 100 km/h with all kinds of different frames and levels of grime on them all moving at different angles.
 

Online xrunner

  • Super Contributor
  • ***
  • Posts: 4783
  • Country: us
  • hp>Agilent>Keysight>?
You could make it recognize a gun or an out stretched arm with a hand gun in it,

Maybe the person is buying a Slim Jim and handing it toward the clerk? That's gonna look a hell of a lot like the end of a pistol.

Quote
or say someone jumping over the counter.

Maybe that's how the clerk gets back behind the counter?

Quote
Even better most employees wear uniforms so it could look for non uniforms behind the counter.

Not universally worn though. Also what happens when the POS (point of sale) technician goes back there to work on the system and he doesn't have a uniform?

Quote
Couple all this with machine learning and I would think its doable.

No - as you can see now it's not reliably possible with today's technology IMHO. You need the perception capabilities of a human level AI. As I said, I believe that will happen, but not anytime soon.
I am a Test Equipment Addict (TEA) - by virtue of this forum signature, I have now faced my addiction
 

Offline X

  • Regular Contributor
  • *
  • Posts: 184
  • Country: 00
    • This is where you end up when you die...
I hope you or anybody else in your life ever has to deal with cancer. I've not been so fortunate and you're a callous dickhead.
You are not special, and you will not gain sympathy from me with name-calling and personal attacks. If you are offended, deal with it yourself. I rarely go out of my way just to offend people, but I am not interested in sugar-coating and censoring myself just to cater to those who might be offended by something I say, especially if it isn't aimed at them.

Are the "black/poor/immigrant neighbourhoods" areas with significantly higher crime rates such because those people are really more prone to be criminals or because those areas are overpoliced, where you can get arrested even for trivial stuff that is ordinarily let slide elsewhere? Especially if you take into account the "broken windows" policing doctrine (= harshly punishing everything, no matter how tiny and insignificant, with the idea that it will discourage more serious crime) that has been widely discredited but is still widely applied.

Now take such data and feed them into a neural network or some other machine learning system. The problem is that if you train that AI with a history of crime that is based on pre-existing biases (immigrant/poor/black/muslim/whatever ...), you will actually only aggravate the situation. Basically a machine-reinforced confirmation bias.

http://www.sciencemag.org/news/2017/04/even-artificial-intelligence-can-acquire-biases-against-race-and-gender
https://www.propublica.org/article/machine-bias-risk-assessments-in-criminal-sentencing
https://www.nytimes.com/2016/06/26/opinion/sunday/artificial-intelligences-white-guy-problem.html?_r=0

It actually gets worse - there are similar software black boxes that will evaluate your likelihood to re-offend and suggest sentence to the judges. Guess who gets hit with much tougher sentences for the same crimes.
Knowing how police departments typically operate, and having seen the shoddy state of many immigrant/muslim/poor areas, it is very unlikely that any over-policing is responsible for the higher crime rate, so it's more likely that these areas are just more prone to criminal activity. It's not "racism" or "confirmation bias" to acknowledge and utilise facts and correlations such as this, and despite what divisive "minority" groups like BLM want everyone to believe, it in't all the state's fault they are over-represented in crime statistics.

As for harsh punishments, what would you consider to be a minor/trivial crime (other than the obvious ones like smoking pot and saying things that offend others)? An incident can erupt from what is meant to be a response to a minor misdemeanor, but was escalated to the point where the offender now has to be restrained with a large amount of force, or they have a very bad rap. I see nothing wrong with imposing a harsher sentence to someone who is more likely to re-offend, so long as this is done fairly and without unjustified prejudice.

There is a good book on why these black box models where we don't know how the system actually arrives to the result a huge problem:
https://weaponsofmathdestructionbook.com/
And therefore, any such AI/neural network learning system should not be used in legal proceedings against anyone, without the output being scrutinised by appropriate human personnel. If the output is used without enough scrutiny, there is a serious issue with how that jurisdiction operates. I do share your concern to some extent, as I can see the potential for dependence on AI in legal matters to escalate beyond control.

These behavior-based approaches have been widely discredited as an ineffective woodoo, unfortunately they are still being sold by the technology vendors. Such a system would be good to alert the operator that something is going on at best, otherwise it produces way too many false positives.
Sure, I'm not saying they are effective or useful, but that they exist. In reality, high res cameras are mostly useful for identifying people in a crowd as opposed to identifying any behaviour. But as you say, many security tech companies love touting their stuff and pretending that it's useful. It's a major issue in the industry, the "oh everyone else are just dinosaurs, you need OUR tech because it's NEW!" This is what confuses people (including legislators) and results in a mess of others (including those in office) having somewhat of an aversion to security systems. It's a dream for some clients who get their security advise from watching Hollywood crime blockbusters and top-secret agent policing fiction TV shows.

Actually most applications of CCTV are for ex-post investigations - once something happens the footage is actually reviewed.
Yes, video analytics are still used in this case, to search for the object or evidence one may be looking for. Even simple analytics such as object searches based on visible movement.

And re behavior based stuff:

http://www.law.uchicago.edu/files/files/Harcourt%20OpEd%20Behavioral%20Profiling%20Longer%20Version.pdf
https://www.rt.com/usa/tsa-behavior-profiling-program-685/

And that is done by humans - it is very unlikely that some sort of AI would be performing any better than that if even humans have difficulty to predict the problem correctly than a random chance.
Of course you still need a human, all the video analytics AI does is make the job easier. There will be false alarms present that the humans have to be around to filter. I am not suggesting the AI should be the be-all and end-all in video analytics, but it can reduce the time an investigation takes.

In most applications of CCTV, such advanced behaviour-identifying analytics aren't even necessary, as the system is composed of many "low" resolution cameras in specific positions, and the location and rough time and location of occurrence is often already known. The only thing the video analytics does here is help the operator find what he is looking for without spending a lot of time sifting through footage manually.

It is concerning how advanced this can get, knowing that the dependence on technology will increase, and even failure modes that are supposed to be non-essential will result in significant losses.
« Last Edit: May 11, 2017, 03:37:01 am by X »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf