Catch up on stories from the past week (and beyond) at the Slashdot story archive


Forgot your password?

Skulls Gain Virtual Faces 279

rw2 writes "Totally cool, The guys at Max Planck Institute for Computer Science have developed a way to reconstruct a persons appearence when a skull is found. When police find a skull and want to know what its owner looked like, they generally use artists who reconstruct the face by building up layers of clay over the skull."
This discussion has been archived. No new comments can be posted.

Skulls Gain Virtual Faces

Comments Filter:
  • Oooh! (Score:5, Funny)

    by bo0ork ( 698470 ) on Thursday August 21, 2003 @04:37PM (#6758546)
    I can't wait to see what that skeleton that hangs in the biology class lab looked like when it was alive!
    • Re:Oooh! (Score:2, Funny)

      by Anonymous Coward
      probably plastic.
      • Re:Oooh! (Score:4, Funny)

        by Anonymous Coward on Thursday August 21, 2003 @04:40PM (#6758599)
        so.. Michael Jackson then?
        • Re:Oooh! (Score:2, Funny)

          by Anonymous Coward
          No, the one in my class had a normal looking nose.
        • Re:Oooh! (Score:4, Interesting)

          by theedge318 ( 622114 ) on Thursday August 21, 2003 @07:26PM (#6760153)
          You should have watched the Nefertiti Special that was on the (Discovery Channel/TLC ?). It was very cool ... this one Egyptologist that specializes in wigs, saw a wig in a museum ... figured out the time period/gender/social status and surmised that it could have been Nefertiti's. She then got permission to enter the tomb where it was found.

          The long and the short ... the show was a bit drawn out ... but they x-rayed the skeleton, shipped it off to a school in England (Nottingham I believe) ... where they blindly (with no a priori knowledge that they would be comparing it against Nefertiti's statue) reconstructed the face from the X-rays.

          The end result was suprisingly close ... especially when you consider that the statue is an artists rendition.

          What really annoyed me was that the producers of the show did a side by side of the CG head and the statue ... and they rotated them at different speeds ... so I had to use the my homebrew PVR just to pause it when the two heads lined up.
    • Re:Oooh! (Score:5, Funny)

      by Captain Large Face ( 559804 ) on Thursday August 21, 2003 @05:09PM (#6758952) Homepage
      I can't wait to see what Skeletor [] looks like. But I already have my suspicions []...
      • Actually, you get to see what he looks like in the first episode of the new He-Man series on Cartoon Network. However, I'm sure that many people will say "the new series isn't canon," or some such shit. I say, just shut the fuck up and watch the cartoons.
      • Re:Oooh! (Score:2, Funny)

        by PrImED73 ( 695394 )
        Actually, its already been reported to be Victoria Beckham
    • Gulp! So that's where uncle Frank has gone!
  • Pretty neat (Score:5, Interesting)

    by stratjakt ( 596332 ) on Thursday August 21, 2003 @04:37PM (#6758556) Journal
    They've been doing this on every discovery channel special on mummies I've seen for the last year.

    Most recently the Nefertiti one that I watched just the other night.

  • Soviet Mobs? (Score:4, Interesting)

    by SHEENmaster ( 581283 ) <travis AT utk DOT edu> on Thursday August 21, 2003 @04:38PM (#6758564) Homepage Journal
    Doesn't the Russian Mafia use base solutions to desolve "enemies", letting their flesh run down the drain, leaving only bones?

    The real reason is to identify McBride's remains after his speech at Defcon.
  • article on Google (Score:2, Informative)

    by kaan ( 88626 )
    read it here [] from the Google cache
  • by RobertB-DC ( 622190 ) * on Thursday August 21, 2003 @04:39PM (#6758580) Homepage Journal
    Interesting article, but just this weekend I watched a special on the Discovery Channel that included this very technique. The cable channel's Nefertiti Resurrected [] special climaxed with a computer-generated rendering of the "mystery" mummy's face, based on the skull and average tissue thickness at key points. They even noted that the technique was "much faster than traditional clay-sculpture reconstruction"... just like the referenced article.

    Jump here [] to see the results.

    By the way, I recommend watching the show. Call me superficial, but I liked the look of the actress who played the doomed queen -- especially her dark skin and freckles. Egypt gets a lot of sun, and SPF 45 was still about 2,900 years away. Much more convincing than Yul Brenner [], and a darn sight better looking.
    • by stratjakt ( 596332 ) on Thursday August 21, 2003 @04:46PM (#6758685) Journal
      The reconstruction scenes were silly!

      Do you think Nefertiti and Akenaten sat silently next to each other, slowly turning at regular intervals to give each other shifty eyed knowing glances?

      And did you see that one priest dude with the leapord skin shawl and the GIGANTANORMOUS AFRO! The Afro was bigger than him! He was the pimpinest ancient egyptian I ever did see.

      And I liked that the whole conclusion that they had found nefertiti was based on "If its not Nefertiti, who else could it be?" Gee I dunno, maybe one of the other BAJILLION people who lived in egypt?

      Anyways. I cant stay mad at TV. I just wish they'd stick to CGI dinosaurs.
      • That afro was fucking obscene. I want to know how they know all that shit. I kinda felt like I was watching futurama.
        • That afro was ... obscene. I want to know how they know all that ... I kinda felt like I was watching futurama.

          Futurama? That looked more like something out of Superfly! []

          Disclaimer: The author of this comment has never actually seen the film, "Superfly".
          • Oh I just meant how on Futurama they do those tours that tell how life was in 2000 or whatever, and their explanation of things is so stupid(the show is brilliant; they are making fun of the way people explain things in the past). Fuck I can't figure out how to say what I am trying to say.
      • And I liked that the whole conclusion that they had found nefertiti was based on "If its not Nefertiti, who else could it be?" Gee I dunno, maybe one of the other BAJILLION people who lived in egypt?

        Well, she was female, aged from 25-30, and must have been a Pharaoh because her right arm was bent across her chest. Mere queens or other royalty bend the left arm (or not at all). That was the clincher. Guess you weren't paying attention.

        If it wasn't her, then there's some other female Pharaoh we haven't

    • by penguin7of9 ( 697383 ) on Thursday August 21, 2003 @05:05PM (#6758921)
      I liked the look of the actress who played the doomed queen [...] Much more convincing than Yul Brenner, and a darn sight better looking.

      Well, unless Nefertiti was a drag queen, it is perhaps not all that surprising that Yul Brynner [] didn't make a convincing Nefertiti.
    • I didn't realize that they had lip augmentation back then.
  • The missing pieces (Score:2, Insightful)

    by dlosey ( 688472 )
    I wonder what it does if a part of the skull is missing. I bet that in many cases, if a skull is found by the police it was a murder. How would the software handle a bullet hole or if part of the skull was crushed. I didn't see it mentioned in the article.

    Could be pretty interesting if there was an extra hole in the face and it put the eye in the wrong spot, or even added an extra one.
    • "handle a bullet hole or if part of the skull was crushed"

      The same way a clay reconstruction modeler does - they extrapolate from the existing bits if it's not possible to glue them back together.

  • Now THAT's useful! (Score:2, Insightful)

    by SoTuA ( 683507 )

    Imagine all that clay savings! w00t!

    Of course, maybe the forensics experts will miss playing with clay...

    For archeology, it sounds cool. Will it work on older skulls, or is it homo sapiens only?

    (tried RTFA... timeout! slashdotted already?)

  • Slashdotted... (Score:2, Informative)

    by rhexx ( 515677 )
    Try the google cache. []
  • by dnoyeb ( 547705 ) on Thursday August 21, 2003 @04:41PM (#6758607) Homepage Journal
    This can not be the case. This is getting rediculous.

    Were going to have to start diseminating slashdot stories on a staggered Timezone based schedule.
  • Missing details (Score:5, Insightful)

    by GreenCrackBaby ( 203293 ) on Thursday August 21, 2003 @04:42PM (#6758626) Homepage
    While this is a very cool idea, the article was missing a few details. For example, did they try it out on actual skulls and see how close they came to the former owner of that skull?

    This last little bit of the article doesn't exactly sell this new technology:
    ' The current prototype figures suffer a problem common to computer-generated faces, said Evison "They look ridiculously mannequin-like."'
    • From a few stories ago.

      Tampa Bay Recognition Cameras []

      Maybe they need to rip out all those camera's and replace them with X-ray machines and utilize this software. Not only would we have a person's blue prints, but could make some very nice composites of what these people look like.

      If people did not want to be "shot" they would need to wear full lead suits all over the place.

      Come on, health risks from long term exposure to X-Ray can't be that bad (joke)
    • Where's the freakin tarball? I'm dying to try this out.

  • by Iron Monkey543 ( 676232 ) on Thursday August 21, 2003 @04:43PM (#6758634)
    How do youo know if the person was a bit overweight and had a double chin or big cheeks? I know I looked ALOT different when I gained about 20 pounds and kept it for a few months till I couldn't afford pizza buffets anymore.

    Also, how can a skull help you determine the shape of the person's eyebrows or the shape of their eyes? And they can't use race as a factor because I know alot of caucasians with various eye shapes.
    • Fat leaves grease marks on the rings in your bones. Like in trees. When they cut your bones open, they can see how fat you were in each year by looking at thickness and greasiness of each ring.

      Dunno about eye shapes, though. Good question.
    • How do you know if the person was a bit overweight and had a double chin or big cheeks?

      Do you question that the ppl on the weight loss ads are actually two different people? If someone shaves thier head and/or eyebrows do you fail to recognise them? Ever seen someone you knew wearing a facemask?

      It may not be 100% accurate, but what more can you do with just a skull? I've seen the discovery channel special on this using clay and averages for the sex, race, etc of the remains, and they had a damn good l
    • by LaMuk ( 257751 ) on Thursday August 21, 2003 @05:25PM (#6759106)
      30 years ago when I was an Anthropology major, some of my professors built faces back for the Las Vegas police. Sometimes they would start with a skeleton that had been shattered into small pieces.

      They were very good about telling age, sex, and race.

      They taught us how it was done. Not that I remember much now. But the amount of tissue on the bones is figured out by how thick the bones got a t insertion points. The thicker the bones, the heavier the load.

      Sex is easiest to tell by the pelvic bones, but also can be determined by size and shape of face bones. Size helped determine race. It got a little tricky if the bones were small. Was it because the person was female or Asian?

      Still they were really good at it and their work identified victims of murders.
    • You beat me to the issue I thought of, but not the application: How about they take a CAT scan (or something), extract the skull portion (or even the skull portion plus muscle portion), then generate models of you with x% less body fat? Could be a great motivator...
  • Pre-Human Skulls (Score:2, Interesting)

    by StingRayGun ( 611541 ) *
    Does it work on pre-human skulls? It would be great to see this work on EVE. It might be more accurate then the "artist's renditions."
  • I want one (Score:2, Funny)

    by tlacicer ( 515153 ) *
    how long before the desktop version is out?

    I wonder how you would test it? They should ban this, I mean it might cause people to start killing each other just so they can see if the software really works.
  • Scientists have reconstructed the face of Lucy, famed early human, using this technology. To little surprise, they found her primitive features closely resembled those of homo sapiens SCO executivus, a recent throwback to more primitive cultures that has surfaced in the deserts of Utah.

    • Scientists have reconstructed the face of Lucy, famed early human, using this technology. To little surprise, they found her primitive features closely resembled those of homo sapiens SCO executivus...

      Highly unlikely. It says here [] that Lucy was a hominid and therefore able to walk upright.
      • Yeah. I guess it's hard to describe someone as "upright" who lies so much...

        Of course, it is possible that homo sapiens SCO executivus is the missing link between Lucy and invertebrate animals...
  • i read 'max factor' first - i didnt eat much today and something seems to be wrong with me so it seemed logical they would know how to reconstruct a face being a cosmetics company and all - then it hit me that they wouldnt have an institute for computer science, they usually have some flashy institute for beta caleotones or whatever the latest face goo is called. .. max planck makes a lot more sense in the end than max factor.. maybe i shouldnt post this. or post it anonymously? ohwell
  • by default luser ( 529332 ) on Thursday August 21, 2003 @04:46PM (#6758687) Journal
    Think about the features that people usually associate with a face: eyes, eyebrows, hair, nose, lips...

    All of these features are soft, that is to say that there's very little chance you can extrapolate them from the skull's bone structure.

    Yes, you can get the basic size of the lips and eyes, and the basic width of the nose. But you cannot tell the eye color, or the lip hue, or the actual shape of the nose or eyebrows.

    You would need to extract such things from DNA, if that's even possible today.
    • eyes are not that difficult, they have the size of the eye socket.

      eyebrows are not that important to ID someone, some women go off mucking with them, but I'm sure all of thier loved ones will still recognise them.

      hair changes all the time face is more important.

      nose can be very accurately reproduced from skull measurements and racial/sex averages

      lips too can be reproduces from averages

      eye color, lib hue, etc is not necessary for an ID. B&W photographs have worked well for years.

      If you've never se
  • I remember seeing an episode of Quincy (Jack Klugman was a forensic pathologist) when I was in high school where he reconstructed a guy from a femur, IIRC. I don't think we actually believed that could be done at the time but I do remember afterwards people saying that it was based on fact. This was the early 80's. Now with 3d imaging, and the powerful CPUs we have today, I'm not surprised they can get someone's features from a skull.

  • This isn't the kind of technique you can meaningfully discuss without some sort of reported accuracy. If they take the skull of Don King and produce a computer generated image of Danny Devito, then it's not particularly useful.
  • ...the next time in commiting a major crime, use acid for any leftover evidence... (joking!) :)
  • Jaw bone lifestyle (Score:4, Interesting)

    by swtaarrs ( 640506 ) <<ten.tsacmoc> <ta> <srraatws>> on Thursday August 21, 2003 @04:48PM (#6758717)
    This sounds interesting, but sometimes this reconstruction thing can be taken way too far. I saw a special on either the Discovery Channel or TLC where they found half of a lower jaw bone. From this, they reconstructed the rest of the jaw. Then they reconstructed the rest of the face and head. Then they figured out his eating habits. Then from those eating habits they figured out the whole lifestyle of this guy, from only his jaw bone.... It was interesting but didn't seem very believable.
  • such techniques?

    I've seen this technique used in "found skeletal remains" crime investigations and archeological investigations and have always wondered if the technique was accurate or just being done for dramatic effect.

    Maybe they could dig up a skull of someone who has an available photograph. Give the skull to three "artists" and see how close the results compare.
    • I'm sure they have and that they will work their damnedest to ensure that nobody actually does it. This technique simply doesn't work. Skull shape doesn't determine face shape. Simple as that. It doesn't matter though. The fact that computers with video cameras can't pick out criminals from a crowd never stopped people paying money for systems to do this. The same will happen here.
  • Not Scientific (Score:5, Interesting)

    by lawpoop ( 604919 ) on Thursday August 21, 2003 @04:49PM (#6758731) Homepage Journal
    We covered facial reconstructions in one of my archaeology classes. Basically it's guesswork and artistic interpretation.

    Sure you have the facial bones, but you have no idea how thick their muscles were, how fleshy their skin was, lip size, what their eyebrows were like, eye color, eyelid characteristics.

    There was one study where they gave the same skull model to five different artists and they got back 5 very different heads.

    The only way you could to this accurately would be to decode any DNA you find and grow their face, virtually (or in some vat -- yech). The technology is a long way off, needless to say.

  • CSI?!?! (Score:5, Funny)

    by thebatlab ( 468898 ) on Thursday August 21, 2003 @04:54PM (#6758785)
    This is not new. Doesn't anybody watch CSI? With the aid of computer technology they are able to zoom in on images taken from blurry security cameras to be able to tell if there is a carpet fiber on the jacket of the person in the very same picture! I'm sure they're able to fully rebuild a complete person from just the skeletal structure, muscles and all. They can probably interpolate from marks on the skeleton and thanks to that guy that knows everything he could probably help out b/c chances are he knew the guy. TV wouldn't lie to me!! Would it...?? *cowers in the corner*
  • You could use virtual historical characters from skull reconstructs to play the actual individuals. Getting Henry the Eights head out of the grave might be a pain though. Oh no now we might have a real version of Otsi "or however you spell it" the ICE MAN COMETH! A giant spectacle with European Ice man shorties taking on Arnold the Testosterone Giant. That would be a hoot.
  • So... (Score:5, Funny)

    by docbrown42 ( 535974 ) on Thursday August 21, 2003 @05:00PM (#6758855) Homepage
    ...we will finally be able to see what Calista Flockhart [] REALLY looks like?

  • ...until they can get the flesh color right automatically ;-)
  • In this [] you can see a face reconstructed from a skull, all computer generated, no clay.
  • when MacGuyver did this years ago using eraser heads to build up the thickness of the skin and then recreate the face of a skull he found?
  • Have they reconstructed Otzi the warrior's face yet. Any pictures?

    Cool technology though. I wonder if they could extrapolate to the skeleton maybe by scraping the bones or looking at dna to get a body fat percentage and then get a full body view.

    I wonder DNA analysis could yield body hair, musculature, and other specifics to find a full body picture. Imagine, we might get to see computer generated pr0n of our ancient ancestors. How hot would that be?
  • Reanimating the Dead (Score:4, Interesting)

    by rpiquepa ( 644694 ) on Thursday August 21, 2003 @05:32PM (#6759177) Homepage

    I also covered this subject today on my blog [] where I gave some additional references, including an illustration of a face reconstruction process.

    And remember that this software was shown during last Siggraph. New Scientist published "Animation lets murder victims have final say []" on this work about two weeks ago with a nice illustration, "How the dead can express themselves []."

    In "Skulls gain virtual faces []," Technology Research News didn't give much more information.

  • by Sabu mark ( 205793 ) on Thursday August 21, 2003 @05:42PM (#6759279)
    Maybe I should RTFA, but I doubt that, when the Animal Learning Discovery Travel Court Channel showed a face reconstruction from a skull last week, the method was subjected to SCIENTIFIC SCRUTINY. For instance, judges could compare each CAD face to a series of photos, one of them being an actual photo of the skull model (old family photos could be used if the skull model is deceased) and select their best guess. If the average correct photo cannot be selected by more than N% of the judges, the technique cannot be held to be scientifically valid.

    Why don't people demand this level of veracity from everything in their life? People down herbal placebos by the truckload and spend big bucks for "ancient Chinese traditional medicine" without even realizing or caring that no scientific study has ever verified such practices. People don't even understand what science IS. They think scientific ideas are just one class of things, existing alongside "traditional," "spiritual," or "alternative" theories. This is ludicrous. There are only two categories of things - things that truly exist or truly work, and things that don't. And the only reliable way to tell them apart is through the scientific method, not an appeal to the supernatural or something's ancientness. How can people have been so inadequately educated? Ugh! I hate everybody.

    Sorry, my misanthropy flared up again (as I have trained it to). But on a related note, the Animal Learning Discovery Travel Court Channel also has lots of other forensics shows where they show hair analysis and "blood spatter analysis." And I want to know whether ANY of these things have ever been scientifically established, or whether (and this is my suspicion) they're partially or totally bogus but more than convincing enough to fool the average jury member - who himself probably wears an energy crystal and watches John Edward every week. I'm skeptical about even fingerprint analysis. Has there ever been a study done to support them? I don't know. Every schoolboy is taught about fingerprints and how each one is unique, but what if their effectiveness is just an urban legend that even law enforcement believes? After all, every schoolboy knows about lie detectors too, and those are notorious for being totally bogus, completely unable to withstand and kind of scientific scrutiny. Polygraphs aren't even allowed as evidence. (But, of course, the federal government still uses them for hiring - further proof that the government is stupider even than the average fool.)

    I just hope I'm never accused of a crime. Who knows what kind of "analysis" they'll have come up with. "My office analyzed the victim's facial muscles using muscular memory analysis, and I can say with 99.999847% certainty that the last words formed by her mouth were 'No!' followed by the defendant's name."
  • I've never seen any photos of someone in real-life and then the reconstructed version. This would be possible with a few pictures of those that have died recently. Just find the family and get a picture of them after the software has done it's thing.

    I'd like to see just how close they come to actually getting it right.
  • Actually, this method (3 dimensional reconstruction of musculature and flesh upon skulls) has been used within anthropology. Here's a site with some interesting photos and explanations [] of the process used. Pretty informational. Enjoy.

Perfection is acheived only on the point of collapse. - C. N. Parkinson