Chicago Robber Caught By Facial Recognition Sentenced To 22 Years 143
mpicpp (3454017) writes with this excerpt from Ars: "The first man to be arrested in Chicago based on facial recognition analysis was sentenced last week to 22 years in prison for armed robbery. ... In February 2013, Pierre Martin robbed a man at gunpoint while on a Chicago Transit Authority (CTA) train. After taking the man's phone, Martin jumped off the train. However, his image was captured by CTA surveillance cameras and was then compared to the Chicago Police Department's database of 4.5 million criminal booking images. Martin, who already had priors, had a mugshot in the database. He was later positively identified by witnesses. At trial, Martin also admitted to committing a similar robbery also on the Pink Line in January 2013—his face was captured during both robberies."
Re: (Score:2)
"This is the guy our fingerprint comparison said did it, does he look like the guy?"
Re: (Score:3)
That isn't how lineups are done in real life. Real police work bears no resemblance whatsoever to the routine felonies committed by character of cop shows.
Re: (Score:2)
Agreed. Which is why the grandparent seems off base with that comment. The police are not going to be feeding hints to a witness during a line up.
It's all irrelevant to this particular case anyway, and I think people are instantly imaginging worst case scenarios for facial recognition rather than reading the story.
Re: (Score:2, Troll)
The police are not going to be feeding hints to a witness during a line up.
Hahahahahahahahahahahahahah. Oh wait, were you serious?
At this point, why would anyone give any cop the benefit of the doubt about anything?
Re: (Score:1)
Re: (Score:2, Informative)
Any remotely competent lawyer would get that kind of identification thrown out of court. Any lineup, even a photo lineup, without multiple options is inadmissible in court.
Re:My two cents (Score:5, Insightful)
Sounds to me like this was used as an investigative lead that helped them find other evidence, rather than as the principal evidence presented in court. This really isn't different than a police officer viewing the recording to see the offender's face, then going through books of mugshots to find the face, then investigating those people that the officer thinks might be the offender. This is simply the computer taking the image that the police officer identified and searching those "books" for close matches, then the police looking at the MO of the crime as compared to the MO of the person previously arrested, and investigating ones that have the most commonality first.
In this case they identified a suspect, the suspect apparently had offended in this same way before, and the suspect was tried and convicted. This doesn't seem to violate any new privacy considerations- the recordings being collected themselves are nothing new, and the mugshot database isn't either. Simply making the comparison itself doesn't add any new fuel to the fire of personal liberty complaints or of violation of privacy.
Re: (Score:2)
Imagine this scenario: I don't know if this person did it, but if the facial recognition software says it's true, it must be him. "Yes, officer, that's the guy."
I was responding to the OP who was implying that the officer would only show the witness one photo and stating that the facial recognition picked him.
Re: (Score:2)
Besides, based on the TV "police procedurals" that have been on for the last fifteen years, I expect that a statistically significant portion of the population already believe that this sort of facial recognition was already going on. Given that I remember ac
Re: (Score:2)
I'm all in favor giving people opportunity to change their behavior for the better, but I'm not going to
Re: (Score:2)
This really isn't different than a police officer viewing the recording to see the offender's face, then going through books of mugshots to find the face, then investigating those people that the officer thinks might be the offender. This is simply the computer taking the image that the police officer identified and searching those "books" for close matches, then the police looking at the MO of the crime as compared to the MO of the person previously arrested, and investigating ones that have the most commonality first.
Well, since a lot of manual bars were and are lowered when we go from manual matching to computerized search you have to be a bit more careful with that argument. (It's close to being an antique if nothing else).
It's akin to the difference between going out fishing with a pole or two, to scouring the ocean with a fleet of trawlers. In essence it's the same activity, but the effects can be vastly different.
It's for example not at all improbable that the quality of match will decrease significantly when compu
Re: (Score:2)
They don't do that way, that is leading the witness. They include the person they've identified along with other people in a mug shot lineup and ask the witness to pick the person they saw. If the witness picks the same person that the police have already identified, then that is 2 pieces of evidence that they know who did it.
tuttle or buttle? (Score:2)
Imagine this scenario: I don't know if this person did it, but if the facial recognition software says it's true, it must be him. "Yes, officer, that's the guy."
your question reminds me of the movie Brazil. How can someone have done something is the computer says they are dead?
Re: (Score:2)
I would trust facial recognition over humans or a mugshot IF and ONLY IF the recognition algorithm was known and scrutinized/peer reviewed.
A human can make mistakes whereas a facial recognition system can accurately verify individual characteristics and ratios of a face (distance between eyes, ratio of nose distance to mouth etc..
Watch_Dogs (Score:4, Funny)
Re: (Score:2)
Re: (Score:2)
That Chicago is not our Chicago why is the loop an inland? and where is the roads so messed up?
Re: (Score:1)
Fingerprints (Score:5, Informative)
This is nothing more than the type of fingerprint matching that's been going on for many decades. It just puts a name to a person after the fact. Now on the other hand, if he was actively recognized via facial recognition as he was out and about in public and then apprehended, well that would be a different story.
Re: (Score:2, Insightful)
Yet another baseless conspiracy theory from the tinfoil brigade.
Re: (Score:2)
Yet another baseless conspiracy theory from the tinfoil brigade.
Umm, actual reality backs this up pretty well. Look up "parallel construction".
Re: (Score:2)
I doubt they would go through that just to catch a petty thief. By the way I see nothing wrong with concealing sources if the evidence presented in court is legally obtained.
Re: (Score:2)
I doubt they would go through that just to catch a petty thief.
The expression used was "who's to say" that they didn't. It seems pretty unlikely now, but the people who said that the three letter agencies were doing some of the things that we now know for a fact that they have actually been doing were called crazy and paranoid before. After a certain level of complete betrayal, there isn't much reason to give the benefit of the doubt. As for the effort required, who would have thought (except for sane, logical people, who can reasonably extrapolate future trends) that
Re: (Score:2)
Whoops, submitted too soon. I was just going to add about your secret evidence scenario (yes, yes, it's only the provenance of the evidence that's secret, not the evidence itself) that Franz Kafka wrote a lovely little story about just such a court system.
Re: (Score:2)
Who's to say this didn't happen? They match him up with CCTV images from elsewhere. Then, they pretend to recognise him from the train video.
The OP's concern was that they would be matching the face of the criminal from the train footage against faces of innocent people out in public, rather than against mug shots from when criminals were booked. I.e. They'd be doing a dragnet over every face in public, rather than against a collection of faces of known criminals. How does what you're talking about relate to that in any way? I can't imagine a scenario where they'd get any sort of benefit out of doing what you're talking about.
Re:Fingerprints (Score:4, Insightful)
Spell it out for me then, because I'm clearly not getting it.
Near as I can tell, they need to be able to demonstrate in court that they have a way of linking the guy in the train footage to the person they've apprehended. There may be a few links in the chain tying the person to the crime. If the police claim it's via facial recognition from the train footage, they'll need to be able to demonstrate that they can make that identification from the train footage. If CCTV footage gets involved, we've added an extra link to the chain, so they'll need to demonstrate that they can tie the person from the train footage to the CCTV footage (e.g. the person is seen heading in the same direction wearing the same clothes at the same time and location) and then can tie the CCTV footage to the mugshot, otherwise it'll do them no good. And if they're doing that, I don't see why anyone should have any issues with it, since it's no different than going to neighboring stores after a robbery to see if any of them have cameras that got a better view of the suspect's face. That's old-fashioned detective work, not something to fear.
On the other hand, if all they're doing is matching CCTV footage against mugshots, without linking it back to the train footage, then they've failed to tie anyone to anything at all. All they can get from that is "previously arrested person X is currently at location Y", which wouldn't do them much good in court, and it wouldn't be useful to them in the least in getting a conviction since they wouldn't be able to demonstrate the link back to the suspect from the train footage.
And that's before we even begin to address your claims about the NSA stuff, which I find highly unlikely, even with the revelations we've had (everyone knows it's the FBI that keeps the database on US citizens, not the NSA :P).
Re: (Score:2)
They already have a massive database of faces courtesy of Facebook, which they both know about and have easy access to.
FTFY. The NSA doesn't need to steal your info when you've already given it away willingly.
Re: (Score:1)
If they're willing to commit those very serious felonies, then the addition of facial recognition software makes no difference whatsoever. Without it (or, rather, before it), they'd just falsify other evidence instead.
If you believe that all cops are like the ones you see on TV, you should - seriously - move to some place where the nearest other human being is at least 500 miles away. This would be to your benefit, and to everyone else's, as well.
Re: (Score:2)
Processing power would be the main restriction. Running facial recognition on CCTV over a large number of cameras in real time is impractical. Now if you know where he's likely to be or are running it after the fact it's a different story.
Re: (Score:2)
An hour to go through effectively still images looking for numbers using the computing power of Google compared to real time analysis of video for faces using the computing power of a city police department...
Re:Fingerprints (Score:5, Interesting)
I agree, but I think there's another concern here as well: false positives are significantly more dangerous than with other fingerprinting techniques. If DNA samples or fingerprints provide false positives, we have (admittedly error-prone) eyewitnesses as a final layer of defense, and since people who look entirely different can have similar fingerprints or DNA signatures, it's likely that the people look nothing alike. Not so with facial recognition, since a false positive is likely to be close enough to a true positive that it will be incorrectly affirmed by eyewitnesses, even if the authorities don't bias them by telling them that the guy was a match.
None of which is to say that I think we should stop using it, since it is a valuable tool. I merely think that it needs to be used with an understanding of its faults and taken with the grain of salt it deserves.
Re: (Score:2)
since a false positive is likely to be close enough to a true positive that it will be incorrectly affirmed by eyewitnesses, even if the authorities don't bias them by telling them that the guy was a match.
That's exactly what I thought when I read in TFA that "he ranked No. 1 among possible matches."
It's a matter of when, not if, the #1 match is innocent, but was in the same place at the same time as the actual perpetrator.
Re: (Score:2)
It's a matter of when, not if, the #1 match is innocent, but was in the same place at the same time as the actual perpetrator.
So take away facial recognition and what changes?
Basic old-school law enforcement:
step 1: get a description of the perp from witnesses
step 2: get a list of suspects -- find out who was there
step 3: show the witnesses the suspects and see if they recognize the guy
If you were at the scene of the crime, and looked like the perp, odds are decent you are going to get busted. After all, th
Re: (Score:1)
Facial recognition isn't necessarily even going to make it worse.
It might put a lot of extremely similar looking people into a line up and that cant be good.
Re: (Score:2)
Generally speaking, there is an attempt made (as there should be) for all the guys in a lineup to look similar, which means your argument is again all lineups, despite that being proven the best way to do such things.
Re: (Score:1)
Hopefully there's other evidence, and video surveillance might show clothing the person was wearing which he or she might still own. Worrying about edge cases isn't necessarily arguing against lineups. Just making sure they're as foolproof as possible.
Re: (Score:2)
which means your argument is again all lineups
Not really.
In a traditional lineup, the police will have identified the suspect using an independent factor (e.g. seen at the time and place, crime fits their M.O., DNA evidence, they were later heard bragging about the crime, etc.), and the eyewitness is demonstrating their reliability by picking out the suspect from among people that look roughly similar. When the eyewitness identifies the suspect in the lineup, the police have now identified the suspect based on two (or more), independent factors that re
Re: (Score:2)
Why do you think DNA samples or fingerprints are more likely to have false positives than (you admitted) very poor human memory?
I'm not claiming the tech is always better, but at least with DNA samples, and I am under the impression with fingerprints (please disprove my belief), they at least have reasonable stats at how likely it is to have false positives... as opposed to "a 6 foot tall guy with blond hair".
Re: (Score:3)
I think his point is that fingerprint and DNA false positives dont lead to a suspect that looks like what a witness saw. Whereas facial regonition false positives almost guarantee that the person will at least look similar to what the witness saw. Thus for facial recognition, the witness-as-a-confirmation is not as compelling. It's almost the same piece of evidence, rather than two corroborating pieces.
Re: (Score:2)
I think his point is that fingerprint and DNA false positives dont lead to a suspect that looks like what a witness saw. Whereas facial regonition false positives almost guarantee that the person will at least look similar to what the witness saw. Thus for facial recognition, the witness-as-a-confirmation is not as compelling. It's almost the same piece of evidence, rather than two corroborating pieces.
That's a very good point, and well worth considering, especially given the now known fallibility of eyewitness accounts. (Not that courts want to really consider that, since that would make convicting someone much, much harder.
On the flip side. This match is one which humans are well equipped to reason about. We know instinctively what "likeness" means and it's easy for (almost) everybody involved to judge the similarity between i.e. a mugshot and a grainy surveillance video. In fact the quality of the evid
Re: (Score:2)
This post [slashdot.org] by alostpacket accurately sums up what I was getting at. I never claimed that fingerprinting techniques produce more false positives than eyewitness identification (in fact, I believe the opposite to be true). Rather, I was pointing out that DNA/fingerprints are independent of how an eyewitness identifies someone, so we can rely on eyewitness identification as an independent factor by which we can verify those earlier tests and hopefully root out any false positives. Not so with facial recognition
Re: (Score:1)
I'd imagine if the only evidence was a photo then it would be thrown out as circumstantial. People look a lot more alike than fingerprints.
FTFY (Score:5, Insightful)
Chicago Robber Identified By Facial Recognition Sentenced To 22 Years
Caught would imply that he was walking down the street and facial recognition directed authorities to him. That did not happen.
Re: (Score:2)
Chicago Robber Identified By Facial Recognition Sentenced To 22 Years
Caught would imply that he was walking down the street and facial recognition directed authorities to him. That did not happen.
Police state would imply they're always watching you, whether they arrest you on the spot or come by later. There's also no real line for the police to cross except better technology and that will come.
Re:FTFY (Score:4, Insightful)
Wow, somebody being arrested for an actual crime that the suspect actually committed is a "police state"? In a public place it is best to assume someone is always recording so don't commit a crime.
Re: (Score:2)
I am not concerned about this crime but rather how this technology can and will be used. I suppose one could argue this is no different than using fingerprints to catch a crook, except it is vastly more than that. AFIS only contains a small portion of the U.S. population’s fingerprints, mostly those who have already committed a crime. I doubt who decides everyone should be forced to give up their fingerprints and DNA while they’re at it to complete the database would have his job very long to
Re: (Score:2)
You missed one huge point. His face was matched against an arrest record. Finger prints were taken during the booking and so was his photo. It is exactly like fingerprints in this case.
Because far too many have already surrendered to the idea that “public” space means the government can watch you,
There was nothing to surrender. The government could always watch you in a public place.
An image with a likeness and couple of witnesses who agree it looks like him is far more tangible to a jury
I guess you don't understand the rules around a photo lineup [innocenceproject.org]. A photo lineup done wrong can get thrown out of court along will all evidence subsequently found.
When combined with data mining, the government will have the perfect capability to track and essentially know all peoples movements, anywhere, anytime.
Sorry but "Person of Interest" [wikipedia.org] is not reality and won't be for quite some time.
Re: (Score:2)
If we could limit photo matches to just arrest records, that would be one thing but although I don't have time to look up a citation, it's also being done against drivers license photos and it's not hard to see it extending out from there. Also, I never said the technology to do real time scans was available today, only that it will be in the near future. We also don't have anywhere near complete camera coverage but you don't even need anywhere near 100% to make life oppressive. And yes, there are evidenti
Its, ... its, ... (Score:2)
Re: (Score:2)
Wearing a mask is illegal in many states unless for medical reasons or weather.
http://www.anapsid.org/cnd/mcs... [anapsid.org]
Re: (Score:3)
Wearing a mask is illegal in many states unless for medical reasons or weather.
Your own source seems to disagree with you. According to it, about half of the states blacklist specific, prohibited activities, but otherwise allow masks for anything else, while the other half whitelist a broad set of permitted activities that hit most of the common cases, but otherwise disallow masks.
Among those that blacklist activities, the lists are pretty much all the same: no wearing masks to conceal your identity while engaging in crime (i.e. it's one more charge they can add on top), no wearing ma
Mirror, seriously (Score:3)
No Dude, poor life choises put you behind bars, the best years of your life down the tubes for a smartphone. This is a perfect example of how stupid is a action verb, not a state of being.
Re: (Score:2)
Actually, more like "poor birth choices" in the united states.
15% of the top 1% are children of the top 1% 20 years ago.
30% of the bottom 20% are children of the bottom 20% 20 years ago.
Interestingly, a portion of the top 1% also flips back and forth between being in the top 1% and a negative income or zero income.
If you are born poor, educated by substandard schools, lack a stable family- your odds of "making poor life choices" is much higher.
Re: (Score:1)
Re: (Score:1)
As a criminal defense attorney who has represented many hundreds of defendants (including a large number of indigents), I regret to inform you that this is not actually true.
The average criminal case usually involves a mountain of evidence left behind a poorly planned and executed crime. The defendant's guilt is painfully obvious in 90 percent of cases and as a defense attorney, all you can do is try to maneuver your client into a less crappy bargaining position so he gets a palatable plea offer.
Actually "b
Re: (Score:2)
The top 1% is defined by annual income tax filings. This is the common usage. It's what is being referred to when they say the top 1% has turnover and changes.
An alternative top 1% is to break it out by holdings. That's not commonly used. I suspect it's a more stable indicator than income but I bet it's damn hard to actually calculate since many of the assets value aren't set until they are sold.
Re: (Score:2)
A large percentage of the 1%'s only qualification for being in the top 1% is that their parents were in the top 1%. It's about 17% from what I recall.
An even larger percentage of the 1%'s only qualification is that their parents were in the top 10% (about 45%).
There are MANY rich and powerful idiots.
Some PD's are as you describe. And (having served on multiple juries) some really don't give a shit and just phone it in.
In one case I was in- it was hard to tell who was more apathetic- the PD or the substitu
Return on Investment? (Score:1)
It's kind of like the T. S.A.: jillions spent to catch one guy every 3 years.
Re: (Score:2)
Re: (Score:2)
"I forgot I was carrying it" is simply the most common excuse given by people caught with them, and is not necessarily the actual reason.
Re: (Score:1)
TFA doesn't have his face... (Score:1)
So why the heck can't they show his face in a story about facial recognition? Why the picture of a train? That has nothing to do with facial recognition! For all we know he has some incredibly unique face or maybe a tattoo across his forehead.
Re: (Score:2)
So why the heck can't they show his face in a story about facial recognition? Why the picture of a train? That has nothing to do with facial recognition! For all we know he has some incredibly unique face or maybe a tattoo across his forehead.
There's 2 links in the summary - not to mention plenty of other articles about this exact story - the second one [suntimes.com] includes a photo.
Why a train? Probably because it was about a robbery that occurred on a train, but why are you asking that here when you could ask the author [arstechnica.com]?
Coming soon... (Score:2)
Mugshots of everyone so they don't have to wait for priors to be able to use this technology.
Oh, wait. They're already half-way there with state IDs.
Re: (Score:2)
Have a passport? You already in the database. Served in the military or ever been fingerprinted? You already there as well.
Re: (Score:2)
Most jobs of any significance require you to give your finger prints up anyway. Not a big deal if they want to put my ID photo in the big computer as well.
Re: (Score:2)
Facial Recog has a high failure rate (Score:2)
Just saying.
All this will do is put stupid people in jail, while high-stealing bank execs walk the streets free.
Re: (Score:2)
The best way to rob a bank is to own it...
Re: (Score:2)
How to rob a bank [ted.com] ... /citation
Re: (Score:2)
Actually:
http://www.amazon.com/The-Best... [amazon.com]
Re: (Score:2)
Blanket statement that just isn't true.. nice try tho. With a high quality gallery (which the mugshot gallery is) you can obtain failure rates significantly less than 1%. The big question here is the quality of the sample taken from the CTA's cameras. Angle and Resolution are the biggest issues with CCTV footage but quantity of cameras and availability of low cost/high-rez equipment are rapidly eliminating both.
Having spent years deploying these systems I'm sorry but your claim is just plain false. The
Wear a balaclava (Score:4, Interesting)
But it doesn't justify the mass surveillance being put in all over our public spaces. It can't even be justified on the cost, but far worse is the erosion of your freedom to go about your business without being tracked and monitored permanently. It might catch the odd transgressor, but that is not an acceptable enough reason to piss away all our privacy.
Oh but you have nothing to hide, so what? Well, it was Joseph Goebbels who first made that pithy remark about having nothing to fear, and look where that ended up - many perfectly innocent people had everything to fear.
The only reasonable response to mass CCTV is for everyone to wear a balaclava. Once the system is rendered useless, they might reconsider spending taxpayer's money on it. And it sends a strong message that we simply don't want to be tracked, even if we are not criminals.
Re: (Score:2)
Criminals doing crimes in masks... work with me, people....
Re: (Score:1)
Re: (Score:2)
The only reasonable response to mass CCTV is for everyone to wear a balaclava.
That's silly.
People willing to wear balaclavas to avoid being tracked are clearly willing to take the much less radical step of voting for and donating to the campaigns of politicians who oppose mass CCTV coverage. If you get a sufficiently large segment of the population willing to wear balaclavas that the CCTV system is useless, then you also have enough public opposition to CCTV cameras to remove them via the political process.
The truth of the matter is that most of the population doesn't care, and a
What"s A Criminal To Do? (Score:2)
well under the GOP system better doctor then ER (Score:2)
well under the GOP system better doctor then ER and you get stuff that the ER does not do.
Re: (Score:3)
Crime is no longer a career choice.
Armed robbery of people on a train haven't been a profitable profession for at least 150 years :) :)
And I'm basing that fact that it ever as profitable on movies
Crime has long been the employment of quite a few members of society but now they will be caught.
s/employment/desperate measure/
By the way, criminals being caught is not a new thing... close to 1 percent of the prison service eligible US population is behind bars.
Re: (Score:1)
Re: (Score:2)
Re: (Score:2)
Any idiot criminal can evade this system. He'll just bring a mask hidden on him, enter a crowded bathroom, put on the mask, steal money and leave.
This is just PR crap for a big brother system that will track where everyone is at any given time in public.
Booking Photos Database (Score:2)
Re: (Score:2)
What's the procedure about booking photos (and fingerprints taken at booking) in the US? Is it possible that your image could be on that database even if you were not convicted of a crime...
You're booked when you're arrested, which is long before your trial. So lots of people have had mugshots taken who later were exonerated.
I doubt they are going to thrown out perfectly good records once they have them.
Re: (Score:2)
Since When... (Score:1)
Since when do eyewitnesses "positively" identify subjects?
Heavy hand of the Law (Score:1)
Is it just me or does the sentence sound unusually extreme, I mean using a firearm in the robbery should probably add to the sentence in most countries, but 22 years for stealing a phone and perhaps something similar one other time seems disproportionate.
The computer says it so it must be true (Score:1)
I finally got screwed by ebay/paypal this year. Bought some cables to connect up some solar panels and the seller gave a tracking number that said delivered, even though I was home all that day, have video of the package not being delivered, but too bad so sad you are SOL.
So I lost $130, but just imagine what happens when the software says you did it when you didn't.
Finding the wrong match (Score:2)
Basically this system is going to be excellent at finding both the correct people and their doppelgängers. I certainly hope that in this case they were able to find some solid evidence.
But if they extended their database search a bit further into
Slippery slope (Score:2)
Everything in the world is a double-edged sword. Another example is DNA evidence.
For over a century, fingerprints have been the gold standard by which suspects were positively identified. Today, the reliability and uniqueness of an
Re: (Score:1)
No one piece of evidence alone is going to convict.
22 years for theft, even with a gun, is harsh (Score:1)
Re: (Score:2)
Obviously this criminal's previous arrests didn't serve as a wake up call. This sentence quite appropriately protects the public -- specifically the next group of innocents that this guy decides to rob at gunpoint. It also teaches the guy's children that crimes have consequences.
Re: (Score:1)
Re: (Score:2)
Most criminals are black. That's more than just a coincidence.
Worldwide? I would guess that most criminals are Asian in "ethnicity".
Re: (Score:2)
Shhh... Slashdot is not the place for common sense.