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

 


Forgot your password?
Close
typodupeerror
×
AI Science

Computer Spots Fakers Better Than People Do 62

Posted by timothy from the just-don't-make-eye-contact dept.
Rambo Tribble (1273454) writes "Using sophisticated pattern matching software, researchers have had substantially better success with a computer, than was obtained with human subjects, in spotting faked facial expressions of pain. [Original, paywalled article in Current Biology] From the Reuters piece: '... human subjects did no better than chance — about 50 percent ...', 'The computer was right 85 percent of the time.'"
This discussion has been archived. No new comments can be posted.

Computer Spots Fakers Better Than People Do More

Computer Spots Fakers Better Than People Do

Comments Filter:

  • In 3 .. 2 .. 1 (Score:5, Funny)

    by OzPeter ( 195038 ) on Saturday March 22, 2014 @06:00PM (#46553979)

    Outlawed by FIFA!

    • But, but... The referee's humanity is part of the game! It's not right to question his judgement with gadgets!

  • It wasn't the computer (Score:3)

    by fullback ( 968784 ) on Saturday March 22, 2014 @06:02PM (#46553989)

    The people who programmed "the computer" were better.

    • Re: (Score:3, Insightful)

      by Trepidity ( 597 )

      I'd say it's a mixture of the two. The computer can't discriminate these facial features without people to program it, but the people can't discriminate these facial features on their own, either, because we aren't good at applying this kind of analysis ourselves (even if we can come up with what it ought to be). The existence of a computer isn't enough, and the existence of the people is also insufficient, to carry out the task. So I'd call it a collaborative activity.

      • I'd call it people using tools. Only if the computer is intelligent with motivation does it become collaboration.

        • Re: (Score:3)

          by Trepidity ( 597 )

          I'd say the computer is pretty intelligent. For one, it's better at recognizing facial expressions than people are! ;-)

          I mean, if you hired a textile worker, nobody would object if you talked about the worker being "good" or "bad" at sewing, even though they didn't design the sewing machinery and aren't exhibiting any particular creativity, but rather are just following instructions.

          • Re: (Score:1)

            by Anonymous Coward

            You are saying this yourself: the textile worker is neither creative nor intelligent, he is just skilled. What's so difficult to grasp? Can't the computer be skilled at pattern matching? Doesn't mean it's suddenly intelligent.

          • "I'd say the computer is pretty intelligent." Then you are wrong. If turned on, OS allowed to settle and no tasks assigned, the computer would sit for as long as power and hardware allowed and do nothing. That is not intelligence.
      • It's all the computer. The people making/training the system could be only 50% at their pain discrimination, but the computer can be trained up to 85% because it's a neural net doing the computer vision matches, and it gets trained. The computer is shown a billion videos of real pain, and told it is real pain. Then a billion videos of fake pain and told it's fake pain. A human might be able to do it at 85% (or better) if they looked at two billion videos to learn from. We shouldn't expect a human's ins

        • Re:It wasn't the computer (Score:4, Insightful)

          by Artifakt ( 700173 ) on Saturday March 22, 2014 @07:21PM (#46554389)

          There's a good precedent for your argument that this is a question of instinctual skill vrs trained skill, but it doesn't take anything like a billion examples to train a person in the example I'm considering. A very common way to teach health care personnel to recognize Fetal Alchohol Syndrome is to give them an album with several hundred photos of people in various life stages, all suffering from FAS. This method has worked since the time when the photos were black and white, and in fact, using color shots or video footage doesn't seem to have any impact on success or the number of examples needed. Once someone is trained that way, the success percentage is in the very high 90s, and stays that way, at least for a typical crreer. Similar methods are used for other diseases, for example most people have learned to spot Down's syndrome from just a few examples, but where the syndrome produces only some of the usual appearance effects, spotting the 'borderline cases' with high accuracy can be taught this same way, usually taking about 15 minutes.

    • Indeed. It would also be interesting to know how much time and, importantly, iterations it took to get that software to that state. Comparing that program with the run of the mill Joe or Jane isn't really fair. I'd like to see the program compared to doctors or medics.

      • Re:It wasn't the computer (Score:4, Informative)

        by ColdWetDog ( 752185 ) on Saturday March 22, 2014 @06:30PM (#46554147) Homepage

        That would be an interesting test. I agree, 25 random volunteers really isn't all that high of a bar. Do the same with some experienced clinicians and see what happens.

        Moreover, pain is a pretty complex issue - there is acute, nocioceptive (pain receptor) pain - as in the test. Chronic pain is quite a bit different. Visceral pain (from nerve fibers in the abdomen) is different still.

        I think that a computer assisted study of emotions has the potential for improving human performance in decoding those emotions, but this is clearly in it's infancy. I don't think there will be an app for that in the near future (a real one, that is).

    • What did they use to teach the computer?

      Did they torture a bunch of undergraduates with cattle prods, and then load the photographs of the undergraduates screaming in pain into their great big giant neural network of "true positives"?

      Versus photographs of other undergraduates, deeply immersed in their favorite pornography websites, as their "true negatives"*?

      And who were the human beings who were competing against the computers?

      Random joe sixpacks plucked right off of the street, or folks like Milit

      • While you are at it, what happens if they test specifially with medical personnel who have been told they need to spot people faking pain to get their opiate fix,and avoid at all costs encouraging their addiction?

    • It was ***ice water*** not actual pain.

      from TFA:

      In the second, the volunteers immersed an arm in a bucket of frigid ice water for a minute, a genuinely painful experience, and were given no instructions on what to do with their facial expressions.

      For the first one, they told them to ***make a fake painful face***

      I'd say humans did just fine...better than the researchers who designed the study!

      The computer just recognized the patterns it was told...it was optimized on their faces

      another AI hype/fail

    • Re: (Score:2)

      by ras ( 84108 )

      The people who programmed "the computer" were better.

      You don't say why. But I'm guessing if I follow you logic the people who programmed Deep Blue were better at chess than Deep Blue itself, or the people who programmed Watson were better at Jeopardy than Watson. Since computers did these tasks better than the best people in the world clearly they weren't, and by a large margin.

      Technically computers can do some things better than us. For example, they can store a series of images of someone responding to pain perfectly for long periods of time. Humans can

  • Hollywood (Score:5, Interesting)

    by AlphaWolf_HK ( 692722 ) on Saturday March 22, 2014 @06:12PM (#46554059)

    Perhaps watching faked facial expressions on TV and whatnot has dulled our ability to distinguish them?

    • because people never faked emotions or facial expressions beforehand?

      half the population has been faking orgasms since the dawn of humanity

    • Re:Hollywood (Score:4, Interesting)

      by Artifakt ( 700173 ) on Saturday March 22, 2014 @07:34PM (#46554441)

      In the 1970's there was a book called "Four Arguments for the Abolition of Television", or something like that. One of the arguments was the limited image quality of the 512 line scan made even very poorly faked emotions very hard to distinguish from the real thing, and so children who got their learning examples of human expressions from TV would have a hard time telling who was really feeling emotions or just faking them. The author also claimed that emotions such as Rage, Fear, and Strong Suffering would come through better than subtler emotions such as Boredom, Fondness or Compassion, so TV scripts would come to emphasize those emotions which at least somewhat worked and ignore the rest. Perhaps there's something to these ideas.

  • My guess is that this would be useful for detecting people who only want narcotics to sell, or to use recreationally, But a computer algorithm that falsely identifies pain sufferers as fakers would be downright cruel, whereas a computer algorithm that fails to identify fakers is merely less useful for drug control efforts.

    • I always hated how both types of error were given the incredible opaque designations of "Type I" and "Type I"...

    • I often wonder how well our medical establishment has studied the euphoric effect of opiates and how they contribute to or even in some cases surpass the functional pain relief.

      I had a traumatic hand injury two months ago which involved a partial amputation of one of my fingers. I experience a lot of "pins and needles" nerve stimulation and some false limb pain (pressure or stabbing-type sensation where I have no finger) and generalized fatigue in my hand. I take small (5 mg) doses of oxycodone once or t

  • Empathy, inexperience (Score:3)

    by Dan East ( 318230 ) on Saturday March 22, 2014 @06:19PM (#46554099) Journal

    Computers do not feel empathy. Perhaps the empathy humans feel when seeing others in pain overrides any minor inconsistencies in the visual side of things. Further, humans do not "analyze" one others' faces to identify an emotion. We see faces (even if it's the same ones over and over) so many times that it's just an automatic, generalized classification. If people often faked painful facial features, and there was some strong motivation to identify that fact, then I'm sure we would be more adept at it. The only time I can think of in a real-world setting where people fake painful facial features is in jest or to be funny or "sarcastic" in some way (not counting football (aka soccer) matches). Thus the overall context totally reveals the expression to be fake and thus visual side of things is just an afterthought.

    • Re: (Score:2)

      by Jmc23 ( 2353706 )
      Some of us do analyze faces. Especially if the refresh rate on your eyes is fast enough you pick up micro-emotions a lot easier.

      I used to always wonder why people lied so much and why others couldn't tell that they were lying.

      Then I found out how self-involved most people are.

      • Re: (Score:1)

        by Anonymous Coward

        I always find myself wondering why people think I'm lying, even though I'm not. I'm not some liar, and even people I've never met do this.

        So, I'd say humans are pretty bad at telling if others are lying. They get so sickeningly sure that there are arbitrary 'signs' that people give off when they're lying, and, as if not wanting to be wrong, they won't ever admit that they can't magically detect if someone is lying just because they do something. Annoying cretins.

    • another important factor to note: they didn't experience actual pain in the control

      from TFA:

      In the second, the volunteers immersed an arm in a bucket of frigid ice water for a minute, a genuinely painful experience, and were given no instructions on what to do with their facial expressions.

      For the first they told the subjects to make a **fake** painful face.

      So, a face painful face vs a hand in ice water w/ no instruction...vs an optimized facial recognizer optimized on their expressions.

      IMHO the subjects p

  • Still not good enough to easily deny someone pain meds.

    • the junkies will break their own arms to get their fix, so I doubt this method, even if if 100%, would be useful for that. Also, in general, you do not have to demonstrate your pain to get meds. You tell a doctor you are in pain, and he gives you a prescription, 90% of people who get them are not outwardly displaying their discomfort.

    • If you run that software a few times it does. So, someone comes it with fake pains 2 times for meds and gets flagged. There's a 98% chance it's legit. Three times and there's a 99.7% chance they're faking it. You're not going to stop the one-timer, but you'll nail the addict.

  • Banks all over the world uses scanners and computers to spot fake money too, because they don't think the humans can do it as well.

  • I would really want to go over the photos they were showing people. I can think of three different ways this study was contaminated.

    • I would really want to go over the photos they were showing people. I can think of three different ways this study was contaminated.

      because of personal reasons , i am interested in this subject , could you expand on what you have so far contributed PLEASE !

      • Again, its extremely easy to bias a study like this... I'll go through some of my guesses as to how it might be biased off the top of my head in no particular order:

        Looking at pictures of pain in photographs isn't a reasonable test especially if you're doing that to someone cold. That is without a little training. No more then 10 or 20 minutes is what a person needs. But if you just throw it at them out of no where they probably won't adapt to it instantly

        Nearly always when we see people in pain in photos i

  • Comment removed based on user account deletion

  • This technology can get quite dangerous if it becomes good at detecting people with guilty expressions.

  • So random people weren't very good at it, but what about experts in this sort of thing, like a Sherlock Holmes? How would they compare to the computer?

  • due to my having an illness that is at times VERY painfull , and because my GP has known me for over a decade , he does not just look at my FACE ! a couple of years ago , he told his secrety to call an ambulance because of my STANCE , and that was before i even had a chance to say anything to him , due to the fact that i try not to worry my familie , i have learned not to show how much pain i am in !! another thing is that most people do not seem to know that a person in a lot , and i do mean a lot of pain

Slashdot Top Deals

"Protozoa are small, and bacteria are small, but viruses are smaller than the both put together."

Close