First Arrest Made Using Forensic Gait Analysis

rgMekanic

[H]ard|News
Joined
May 13, 2013
Messages
6,943
In a report from Nashville, 3 armed men robbed a liquor store near the Alabama state line in Jan. 2017. Police investigating the robbery had surveillance video, but that showed no identifying features as the suspects were wearing masks, and no idea what vehicle they were in. Police decided to look at video from a few weeks prior to the robbery to see if anyone suspicious was in. One person who was in the store had a distinct walk. The gait of the suspect got the attention of a forensics expert in Indiana, and though nothing but "forensic gait analysis" investigators found the subject who was in the store the day prior to the robbery, was also the robber. Investigators arrested Quinton Nance, who then confessed to the robbery.

Cool story and really incredible article on forensic gait analysis. Since there is now precedence of using gait analysis to get an indictment, I have to wonder if since it's very logical, if machine learning or AI could end up coming to play to automate gait analysis. Thanks to maclem8223 for the story!

District Attorney General Brent Cooper says this is another tool for law enforcement to help prosecute criminals. “The one thing you're not going to be able to disguise is the way you walk,” said Cooper. “We were willing to use this case as a test to see if the courts in Tennessee would accept this kind of proof." This is the first time in the United States that a grand jury indictment was secured strictly using gait forensic analysis.
 
So it sounds like really this only stuck because the guy confessed right? I think a conviction based solely on this would have been pretty unlikely.

Probably but it might have given the LEOs a warrant to investigate the individual further which could have broken the case down the line.
 
So it sounds like really this only stuck because the guy confessed right? I think a conviction based solely on this would have been pretty unlikely.

I have to agree. I reasonable person in the Jury box is presented with video of the suspect from the day before the robbery, and of a "masked" person during the robbery, and told that gait analysis shows that they are the same person .........OK........... And that's all you have and I can have no doubt or I have to vote not-guilty.

With a confession yes, no problems. Without the confession and with no other evidence ..... no way, I don't care how good you claim this technology is, I'd have a hard time backing it.

Now start adding the rest of the physical, same height, approximately the same weight .... you know they cased the store prior because of X and Y ....... it might start adding up so that the aggregate of the circumstantial along with the gait analysis ...... but not on the gait analysis alone.
 
So the moral of the story is, if you are going to commit a crime, then do it with a limp.

Why in the world would the guy confess? Any lawyer worth his salt would have not allowed that. He must not watch enough NCIS.
 
So the moral of the story is, if you are going to commit a crime, then do it with a limp.

Why in the world would the guy confess? Any lawyer worth his salt would have not allowed that. He must not watch enough NCIS.


lol I was thinking the same thing!
 
Why in the world would the guy confess? Any lawyer worth his salt would have not allowed that. He must not watch enough NCIS.
Intelligence? Rather a lack there of, or a nervous "shit! they got me moment"? Had the same thought myself.
 
So the moral of the story is, if you are going to commit a crime, then do it with a limp.

Why in the world would the guy confess? Any lawyer worth his salt would have not allowed that. He must not watch enough NCIS.

There's a whole line of psychology of why people confess, or even give fake confessions. I can't claim to understand it either, but there you go. I guess if you browbeat someone of below-average intelligence enough, state that the video evidence is going to convict them anyway and a confession will give you less time, perhaps you just do it.
 
Here's an idea...don't commit crimes and then you don't have to worry about whether or not the police are looking at gait analysis of the way you walk.

This isn't exactly some white collar crime. They robbed a liquor store at gunpoint.
 
Here's an idea...don't commit crimes and then you don't have to worry about whether or not the police are looking at gait analysis of the way you walk.

This isn't exactly some white collar crime. They robbed a liquor store at gunpoint.

Exactly - which is why when you roll out this (potentially) psuedo-science BS it can be even more damaging, as the defendant doesn't have the cash to pay a real lawyer to argue against it and the jury, fresh off a CSI marathon, convicts knowing if Batman can read fingerprints off bullet fragments, surely the cops can do this. ;)
 
more junk science by law enforcement, just like "bite analysis".

Dystopia, here we come.

While most forensic science is anything but scientific, it can be a tool to help. As long as people understand that it is a tool and not something that can be relied upon as foolproof. Like the whole stupid myth of everyone's fingerprints being different.
 
While most forensic science is anything but scientific, it can be a tool to help. As long as people understand that it is a tool and not something that can be relied upon as foolproof. Like the whole stupid myth of everyone's fingerprints being different.

The average juror isn't intelligent enough to separate pseudo science from reality. All they will know is that mr scienceman is telling them that his fancy AI tells him somebody is guilty, and that's enough for them.
 
The average juror isn't intelligent enough to separate pseudo science from reality. All they will know is that mr scienceman is telling them that his fancy AI tells him somebody is guilty, and that's enough for them.

The average juror is going to look at video footage of the way the guy walks while committing the crime, watch the guy walk, and come to their own conclusion. "mr scienceman" is just going to tell them what they already think they know based on the video. The whole point of "mr scienceman" is giving them some facts and figures to substantiate their judgment.

This is really a borderline, "Nothing to see here" story.
 
Really, I am more inclined to believe the police used parallel construction to identify the suspect and when he was casing his robbery target (same cellphone triangulated to the same location at the time of the crime and the day before the crime occurred).

The gait "evidence" is the ruse to cover for the parallel construction evidence so that the parallel construction evidence which would taint any investigation built upon it (no prior warrants for cellular network data on the suspect prior to using said cellular evidence to track down the suspect) can be excluded from the presented trial evidence and from the official arrest or search warrant requests.

In other words we get to hear about forensic gait analysis as the evidence that broke open the investigation and caught the bad guy rather than the real electronic evidence.

Now, do I know for sure that this particular investigation used parallel construction?
No, I can't prove either way, but I do know unwarranted spying is being used quite often to get the bad guys and parallel construction is the method used to cover that part of the investigation from having to be presented in court a evidence.
There are even cases that have been dropped when the information leaked out that the case presented wad parallel construction and not the real method that caught the suspect.
 
but I do know unwarranted spying is being used quite often to get the bad guys and parallel construction is the method used to cover that part of the investigation from having to be presented in court a evidence.

Quite often? Sounds like a stretch to me, especially with no supporting evidence (and a few examples does not classify as quite often when police investigate thousands of crimes a day).
 
so now criminals walk differently when committing a crime?

criss-cross.

One of the first spy tips I learned as a kid was to put a pebble in your shoe.

This is to slightly alter your gait when you are "on the job" to help disguise your normal movements.
 
Send someone else to scope out the location first. Even movies do that part right.
 
Next up, using gait analysis to predict who is about to rob a store and arrest them before the crime.
 
This really isn't much different from using fingerprints or DNA to identify someone. Its a new method that requires technology to implement it.

I am sure back in the day people were calling fingerprints and DNA as bogus science as well.
 
It's how Micro found The Punisher, so obviously legitimate.
I hope I don't walk like a criminal? I hope I have enough money to fight it? I like the idea of reducing crime, although I wonder how often it could be wrong.
 
So it sounds like really this only stuck because the guy confessed right? I think a conviction based solely on this would have been pretty unlikely.

Yes, but unfortunately they arrested him solely on the basis of this. If you read the story, the analyst called the police and told them that he could "confirm" the identity of the suspect.
 
Yes, but unfortunately they arrested him solely on the basis of this. If you read the story, the analyst called the police and told them that he could "confirm" the identity of the suspect.

Why is this unfortunate? He robbed a store at gunpoint. Blackstone's formulation is all fine and good in theory except when you or your family is the victim of the crime. Remember, the burden of proof is not "beyond any doubt." It is "beyond a reasonable doubt."
 
There's a whole line of psychology of why people confess, or even give fake confessions. I can't claim to understand it either, but there you go. I guess if you browbeat someone of below-average intelligence enough, state that the video evidence is going to convict them anyway and a confession will give you less time, perhaps you just do it.


Or perhaps committing crimes makes some people feel .................... guilty?
 
Exactly - which is why when you roll out this (potentially) psuedo-science BS it can be even more damaging, as the defendant doesn't have the cash to pay a real lawyer to argue against it and the jury, fresh off a CSI marathon, convicts knowing if Batman can read fingerprints off bullet fragments, surely the cops can do this. ;)

You might be amazed what you can get prints off of.
 
The average juror isn't intelligent enough to separate pseudo science from reality. All they will know is that mr scienceman is telling them that his fancy AI tells him somebody is guilty, and that's enough for them.


You don't need the average juror in order for an innocent man to be found innocent, you just need one above average. The beauty of the system my friend, you need 12 to convict.
 
I hope I don't walk like a criminal? I hope I have enough money to fight it? I like the idea of reducing crime, although I wonder how often it could be wrong.

I would add "I hope I don't walk like, talk like, share height/weight proportions, and skin tone of a criminal?" - It's safe to assume they used the gait analysis on video to pick out likely suspects then said "ok, this guy looks just like the suspect."
 
I remember a theory, from years and years ago, that you could uniquely identify individuals based on words they routinely misspelled when typing. This feels a bit like that.

That said, if you feel inclined to commit a crime, you might try some toe shoes (or heels). It'll be the new equivalent of wearing gloves to mask your fingerprints.
 
This ones even better than 'gait'

Cartoonish sketch helps police identify thief

https://www.nbcnews.com/news/us-new...ps-police-identify-thief-pennsylvania-n846336
180209-hung-phuoc-nguyen-comp-njs-1210p_759c18907234dfdf162224a3ac87a214.nbcnews-ux-2880-1000.jpg
 
You don't need the average juror in order for an innocent man to be found innocent, you just need one above average. The beauty of the system my friend, you need 12 to convict.
Not in every US state. There are several people locked up in Lousiana that learned that the hard way.
 
This really isn't much different from using fingerprints or DNA to identify someone. Its a new method that requires technology to implement it.

I am sure back in the day people were calling fingerprints and DNA as bogus science as well.

Plenty of people still think DNA is bogus science. I think we might have one or two in this thread.
 
Back
Top