Slashdot Mirror
Building the Face of a Criminal From DNA
Dave Knott writes: It sounds like science fiction, but revealing the face of a criminal based on their genes may be closer than we think. In a process known as molecular photo fitting, scientists are experimenting with using genetic markers from DNA to build up a picture of an offender's face. Dr. Peter Claes, a medical imaging specialist at the University of Leuven, has amassed a database of faces and corresponding DNA. Armed with this information, he is able to model how a face is constructed based on just 20 genes (this number will soon be expanded to 200). At the moment, police couldn't publish a molecular photo-fit like this and hope to catch a killer. But that's not how Dr. Claes sees the technique being used in a criminal investigation. "If I were to bring this result to an investigator, I wouldn't necessarily give him the image to broadcast. I would talk to him and say okay, you're looking for a woman, with a very specific chin and eyebrow structure."
If you're a bad person we'll reconstruct an identical clone of you and imprison it as your punishment.
If I have been able to see further than others, it is because I bought a pair of binoculars.
How long before a full DNA emulator? Something that would render a fully formed human being starting from his\her full DNA code?
What's the gene that causes high levels of melanin again?
I like how he talks about how he would envision seeing this used, but, I think he actually has it backwards. Not surprizing since, his expertise is in the technique, not necessarily in what it may be used for.
Rather than "you are looking for...." better is to hold this back and narrow down the field. "You have X suspects, now you can eliminate all that don't match this". That will give you better results than "look for people who match this".
This sort of thing has come up many times with the use of this sort of statistic. There are only a handful of blood types, for example, but, if you can say with certainty that the suspect is one of a small group of 2 or 3 people, then blood type might get you down to 1....even though it would be otherwise pretty useless without other information to go on.
"I opened my eyes, and everything went dark again"
While it's a mainstay of many movies from the 1960s and 1970s for criminals to disguise themselves using highly expensive plastic surgery, I'd hardly call it "common", no.
You are not alone. This is not normal. None of this is normal.
I don't even see the code. All I see is blonde, brunette, red-head...
Set your phasers on "funky"!
Everywhere you go in public will be trackable and connectable to your online "public" activities by mapping your DNA to a picture of you online. Laws will not be able to prevent this. This will just become the new norm. . .
.).
Of course, this will also increase the public scrutiny of public officials and other powerful individuals, which I can only see as a good thing (as any "House of Cards" fan should be able to agree with. .
Sdelat' Ameriku velikoy Snova!
While it's a mainstay of many movies from the 1960s and 1970s for criminals to disguise themselves using highly expensive plastic surgery, I'd hardly call it "common", no.
Just grow a beard to hide that distinctively shaped chin, or pluck your eyebrows.
The only thing necessary for evil to triumph is for it to be pitted against a slightly greater evil
Great, now the police have more shaky probable cause to harass citizens with.
Spongebob Squarepants.
Scientist to policeman: My findings in the DNA show the same thing that the witness told us: It was Big Nose Kate.
And what do you do after the next robbery?
Had to say it.
http://www.acetonestudio.com
The output is going to be the "mean" aka the "expectation" but the real world is noisy. You don't just want the most likely estimate, but you want to know the variation around it. What if the ears are really hard to get right, but the philtrum might have exceptionally small uncertainty? Should the police arrest a man for his ears, or his philtrum? The answer is - they must take into account both the ears, the philtrum of the subject, the community, and the model.
This is not a "make a mugshot" until the variability of the model is Cpk 1.3 for the variability of the community in uniquely identifiable features.
This is going to have to work with facial recognition software - so the part that allows one face to be differentiated from another.
Maybe not old as in really old, but at least since 2012/2013/2014.
Even artists know about it.
2012
https://web.archive.org/web/20...
"Researchers have moved one step closer to facial reconstruction with DNA by discovering the genes that help control the width of the human face. A recent study of almost 10,000 individuals revealed five genes associated with different facial shapes – known as PRDM16, PAX3, TP63, C5orf50, and COL17A1. Manfred Kayser and his team of the Erasmus University Medical Center in Rotterdam, the Netherlands, used Magnetic Resonance Imaging (MRI) of people’s heads to map facial landmarks and estimate facial distances."
2013
http://edition.cnn.com/2013/09... "We leave genetic traces of ourselves wherever we go -- in a strand of hair left on the subway or in saliva on the side of a glass at a cafe. So you may want to think twice the next time you spit out your gum or drop a cigarette butt in public. New York artist Heather Dewey-Hagborg might pick it up, extract the DNA and create a 3-D face that could look like you. Her project, "Stranger Visions," fashions portrait sculptures from bits of genetic material collected in public places."
2014
http://www.forbes.com/sites/al...
"Sometime in the future, technicians will go over the scene of the crime. They’ll uncover some DNA evidence and take it to the lab. And when the cops need to get a picture of the suspect, they won’t have to ask eyewitnesses to give descriptions to a sketch artist – they’ll just ask the technicians to get a mugshot from the DNA. That, at least, is the potential of new research being published today in PLOS Genetics. In that paper, a team of scientists describe how they were able to produce crude 3D models of faces extrapolated from a person’s DNA."
http://www.kuleuven.be/english...
"Scientists are getting closer to constructing a likeness of a person's face using nothing but a DNA sample. Postdoctoral researcher Peter Claes and his colleagues describe the technique in a recent publication in PLOS Genetics. Their work opens a horizon of potential future applications in forensics, anthropology and medicine."
Now its 2015.
Criminal investigations is a niche use. The broad use and real revenue would be an adult picture of your unborn child or new baby. Future parents already pay $300-400 for 3d sonograms of their fetus. Imagine seeing your new baby's face each year from two to twenty years old.
You could even sequence a couple individually and show the full range of boy and girl facial outcomes with probabilities. Right now they can use some morphing techniques as a kludge but genetics could be MUCH more predictive.
Creepy, creepy, creepy...
Simple. You get a hair transplant to move those beard hair to your eyebrows.
Old news, been here already
Break my nose and get a big scar on my face. There is no system so perfect someone can't think of breaking.
He seems to say he can do a match on 20 genes.... but that he will be able to soon do over 200 ? So what are those 180 other genes matching for ? Because either what he does today is poor and the 200 are needed or the methodology might even be total crap, or the 180 are not needed if he gets already good result today.
And look at the face it is so generic alone in my building it could match a sizeable part of the population. I would not be surprised to learn that it is snake oil.
But that's not how Dr. Claes sees the technique being used in a criminal investigation.
Yeah, I can see it being used in a lot of ways. A lot of unsavory ways.
-- sudon't
Air-ride Equipped
Digital face construction from DNA is used to identify people for as simple a crime as throwing a used chewing gum on the street.
http://www.digitaljournal.com/...
"We mustn't be caught by surprise by our own advancing technology" -- Aldous Huxley
Does this work on non-criminals?
One very useful outcome is that this could prevent false accusations, false arrests, false inditements and false convictions. The focus tends to be on finding the bad guy but it is even more important to avoid convicting the innocent person. DNA testing does a lot to overturn the convictions of innocent people. This could stop the process of false convictions even earlier.
http://fusion.net/story/154199...
We know that Facebook has a vast facial recognition database so good that it can recognize you when your face is hidden, that the FBI has built a millions-strong criminal facial recognition system, and that Googleâ(TM)s new Photos app is so effective at face recognition that it can identify now-adults in photos from their childhood. But now facial recognition is starting to pop up in weird and unexpected places: at music festivals (to identify criminals); at stadiums (to weed out âoesports troublemakersâoe) and at churches. Yes, churches.
"If any question why we died, Tell them because our fathers lied."
'till those reconstructions start showing that most of the perps are from certain easily identifiable groups...
Anything you'd wish on a clone, think of a twin.
If the system can interface with a database of face data it would still help to reduce the number of possible suspects by a huge amount. Then by a process of elimination and cross referencing with other data, such as phone meta-data, geo-location etc., it could allow for the very rapid identification of the average, impulsive, criminal who's crimes are opportunistic or reactive. The covert collection of DNA from the small list of final suspects would be all it would take to be sure the correct person had been identified before an arrest warrant was executed, catching them completely by surprise. This would leave innocent, face matches completely unaware that they were under scrutiny, which is a good thing as they should not suffer from the psychological, social and potentially financial harm of been wrongly accused. The only problem I see is if they find their DNA profile is retained after the investigation is completed, but even then there is an argument for using their DNA to refine the system as it will then learn to generate better images that reduce the number of false positives.