Stories
Slash Boxes
Comments

News for nerds, stuff that matters

Slashdot Log In

Log In

Create Account  |  Retrieve Password

Device Developed To Help Socially Challenged

Posted by Zonk on Thu Mar 30, 2006 01:55 PM
from the insert-your-own-self-referential-joke-here dept.
An anonymous reader writes "A device from MIT Media Labs that can pick up on people's emotions is being developed to help people with autism relate to those around them. It will alert its autistic user if the person they are talking to starts showing signs of getting bored or annoyed." From the article: "The 'emotional social intelligence prosthetic' device, which El Kaliouby is constructing along with MIT colleagues Rosalind Picard and Alea Teeters, consists of a camera small enough to be pinned to the side of a pair of glasses, connected to a hand-held computer running image recognition software plus software that can read the emotions these images show. If the wearer seems to be failing to engage his or her listener, the software makes the hand-held computer vibrate."
+ -
story
This discussion has been archived. No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
 Full
 Abbreviated
 Hidden
More
Loading... please wait.
  • So Simple? (Score:5, Interesting)

    by JDSalinger (911918) * on Thursday March 30 2006, @01:56PM (#15028268)
    According to TFA, autistic people cannot discern or interpret a bored look on someone's face, but can realize that feeling a vibration in their hand means that someone is bored. Using a camera (to detect boredom) means that the autistic person is looking at the person he is speaking to. It's interesting that a human could receive image data and be unable to remember what it means, but receive touch data and be able to remember its meaning.
    If this interaction is correct, then a big high five to the geniuses that found the vibration communication channel into autistic minds. Of course if this is not the case, how will a vibrator help?
    This sounds like an unlikely solution to me, but I have not studied autism. Perhaps, the importance of this study is not that it will actually help autistic people, but that our face recognition capabilities are getting to the point of being useable in today's society.
    -C
    • " It's interesting that a human could receive image data and be unable to remember what it means, but receive touch data and be able to remember its meaning."

      It's not about 'remembering' what it means. It's about interpreting what it means.

      "This sounds like an unlikely solution to me, but I have not studied autism."

      'Nuff said.
    • Vibrators ALWAYS help. ;)
    • Re:So Simple? (Score:5, Interesting)

      by hackwrench (573697) <hackwrench@hotmail.com> on Thursday March 30 2006, @02:07PM (#15028413) Homepage Journal
      I have Asperger's, similar to Autism. Show me a picture of a bored person. Tell me the person is bored. If you were to ask me what features would indicate to me that person is bord, I'd be hard press to tell you
      Take the picture away. Show me videos of people and then ask me:
      1. Which of those people have elements that are similar to the bored person.
      I'd be hard pressed to answer that as well.
      • by gEvil (beta) (945888) on Thursday March 30 2006, @02:49PM (#15028869)
        *yawn*
      • by gurutc (613652) on Thursday March 30 2006, @02:54PM (#15028935)
        I also have Aspergers, and one of the most painful things about it is that people can't believe it's possible to miss simple and seemingly impossibly obvious clues.

        It's sad that folks don't know enough, yet still comment, to believe that something this simple would be a huge help.

        I want one.
      • Re:So Simple? (Score:4, Informative)

        by Autistic (613287) on Thursday March 30 2006, @03:03PM (#15029012) Journal
        I agree, but I think there is a little more to it.

        I certainly do not see as much from peoples faces as other people can. But I can see a little. I can see stronger emotions than boredom.

        But the other side is knowing that some type of response is necessary and what that response should be. I may see that someone is angry or is sad, but I don't necessarily know what to do about it. I don't know whether to try to approach and help or stand back and wait. Often times, the hesitation of response is seen as lack of understanding.

        So I get accused of not detecting emotion a lot more than is the case. I can see it, but I don't necessarily respond to it in a way that would be expected. I'll do the wrong thing, or if I know that has failed too many times before, I'll do nothing at all.

        • The only way to function with Aspergers is to consciouly learn to read expressions. It is a power tool if you choose to use it that way. But it is the dark side because it keeps you from having real interaction and backfires completely when you tire. And it is a hell of a lot of work. Aspergers sucks
          • I agree with you on all your points. The real power of Asperger's comes in understanding that in all likelyhood, the human brain is nothing more than a very advanced pattern recognition and playback device. The best argument and discussion of this can be found in the book On Intelligence [onintelligence.org], by Jeff Hawkins.

            After much reading of liturature in various fields of artificial intelligence, emergence, chaos theory, network theory, psychology, etc - I have come to believe that recognizing this is paramount to underst

    • Re:So Simple? (Score:4, Insightful)

      by MobyDisk (75490) on Thursday March 30 2006, @02:20PM (#15028550) Homepage
      It's interesting that a human could receive image data and be unable to remember what it means, but receive touch data and be able to remember its meaning.
      You misunderstand autism. This has nothing to do with memory. Autistic people do not have the facial expression recognition algorithms that most humans have. So someone has implemented such an algorithm on a computer, and then the computer tells the autistic person what the expression means.

      If this interaction is correct, then a big high five to the geniuses that found the vibration communication channel into autistic minds. Of course if this is not the case, how will a vibrator help?


      The problem was no how to communicate with autistic minds. The vibration is irrelevant. This would work fine with a light, a sound, or a big glowing sign that says "shut up dummy, you are boring them to death!" The point is that the input is unambiguous. Unfortunately, facial expressions are very ambiguous.
      • Re:So Simple? (Score:5, Informative)

        by Billosaur (927319) * <wgrotherNO@SPAMoptonline.net> on Thursday March 30 2006, @02:30PM (#15028639) Journal
        You misunderstand autism. This has nothing to do with memory. Autistic people do not have the facial expression recognition algorithms that most humans have. So someone has implemented such an algorithm on a computer, and then the computer tells the autistic person what the expression means.

        I worked with the autistic population for about 7 years. I think it has not been established that autistic people lack facial expression recognition algorithms. From what I've seen, they cannot interpret what they see, lacking the ability to integrate facial expressions of others with their own feelings, and use that to create a picture of what someone else is feeling. The autistic individual tends to treat everything as an object, and they can recognize form and substance, but not emotionality. However, they can learn it, given enough conditioning and reinforcement, albeit it is very artificial and prone to error if certain situations occur which were not anticipated. This device may work as an excellent training tool for those who can use it properly, but it won't solve the problem in the long run.

        • by Pfhorrest (545131) on Thursday March 30 2006, @06:43PM (#15030760) Homepage Journal
          The autistic individual tends to treat everything as an object, and they can recognize form and substance, but not emotionality. However, they can learn it, given enough conditioning and reinforcement, albeit it is very artificial and prone to error if certain situations occur which were not anticipated.

          As someone who is naturally very autistic but has learned to understand the neurotypical mindset, I can tell you that this is dead-on.

          The big difference between a neurotypical and an autistic mindset is that autistics see everything literally, as it is, and do not like to jump to conclusions based on insufficient data. (Though we are often very good at pattern recognition and educated guessing, we recognize that these are guesses and don't mistake them for facts). This quickly gives rise to the typical 'defining characteristic' of an autistic personality in not recognizing others' emotional states, because *there is no direct evidence that people other than the observer feel anything*. An individual's only experience of "inner experiences" is their own, and it is by definition impossible to experience another's inner experiences. To the procedurally-oriented autistic mind, this leads to the conclusion that there's no reason to suspect that such "other people's inner experiences" exist. It's an alien concept to the autistic.

          To the neurotypical, certain behaviors exhibited by other people resemble their own behaviors which are triggered, it seems, by "emotions" or "inner experiences", and so the neurotypical jumps to the conclusion that other people have such inner experiences - that there is some "self" or "I" or "ego" or "soul" that is feeling and thinking in there, and not just a bunch of matter that behaves in certain complex ways. I believe this also explains why severely neurotypical people are so prone to religious beliefs in God or gods - if you're already making the jump to ascribe agency (a necessarily undetectable quality) to certain objects we call "people", why not ascribe such agency to other objects or phenomena, or the universe as a whole?

          Now don't get me wrong, I'm not saying that this is a bad thing to do. In fact it's something that borderline autistic cases like your typical geek are often very comfortable with - the anthropomorphization of computer programs that don't "like" each other, or which "fight" over certain resources, or which "talk" to one another. Geeks understand that these aren't literally true descriptions, inasmuch as we are not ascribing inner experience to these programs, but they are very useful, convenient, and accurate shorthand for describing their behaviors. It doesn't take much to realize that talk of other people's thoughts and feelings and inner experiences is really just the same sort of short hand, and that to any given person's honest and literal perspective, all other people really *are* just objects. (Which is not to say that they should be treated unethically or that there is no basis for ethics, but that's a whole other can of worms there).

          And it doesn't take a whole lot more to go ahead and extend this shorthand to other complex systems, or even the universe as a whole; and from that comes a sort of pantheistic view of God. To talk of "God" is just to ascribe agency to the whole universe, a thinking feeling intelligence "behind" it all, the same way that we can ascribe agency to other people. Both of these cases are equally valid or invalid. They're invalid in that neither one is literally true, inasmuch as it's fundamentally impossible to ever have evidence that they are true, and so we have no real reason to ever think that they are true. But they are both valid, inasmuch as the ascription of agency to other people, and understanding the nature of those "agents", is useful for modelling interactions between people (including yourself) which should ultimately be of benefit to the individual using this model; and likewise, the ascription of agency to the universe and the understanding its nature (even in personified terms) p
    • I don't think I have autism, or even aspergers, but I can't usually interpret when a woman is romatically interested in me. It would be really nice to have a device that could do this.
      • Re:So Simple? (Score:4, Insightful)

        by pHatidic (163975) on Thursday March 30 2006, @02:42PM (#15028790) Homepage
        I used to have this problem, but then I just realized I didn't know what to look for. Like me, I'd guess you just have a bad case of Slashdotters. Here are a few signs for starters:

        For girls you are seeing across the room: She plays with her hair, licks her lips, smiles at you, will make eye contact with you.

        For girls you are interacting with: You squeeze her hand and she squeezes back, you ask her a question and she asks you the question back, you touch her arm and she doesn't flinch or move away, she compliments you on anything, you look like you are going to go somewhere and she asks if she can come, she laughs at your jokes, you walk away and she is waiting for you when you return, she is the one to initiate conversation with you, etc.

        • If she laughs when you're not funny -- that one is huge. Or if she ever utters the phrase "you're so funny" (even if you are).

          Think, this device could be used to get basement-dwelling nerds dates -- not just for the autistic crowd!
          • Yes

            But if the hair is standing up on the back of her neck, her teeth are barred, ears back, tail between her legs, then she's probably to be avoided. Don't make eye contact.
        • Re:So Simple? (Score:5, Insightful)

          by gstoddart (321705) on Thursday March 30 2006, @03:39PM (#15029340) Homepage
          I'll take a stab at this ...
          For girls you are seeing across the room: She plays with her hair, licks her lips, smiles at you, will make eye contact with you.

          Can you accurately identify when she is looking at you or someone else? Do you think all such such signals are as overt as licking the lips? The signals can be a lot more muted and ambiguous, so it can be in a gray area where you can miss something subtle -- or hope to see something which isn't there. If she is NOT currently looking at you is that a summary rejection since all interested people are looking? Nor everyone is likely to have contact initiated based on their looks; you as like as not could go completely unnoticed.
          For girls you are interacting with: You squeeze her hand and she squeezes back, you ask her a question and she asks you the question back, you touch her arm and she doesn't flinch or move away, she compliments you on anything, you look like you are going to go somewhere and she asks if she can come, she laughs at your jokes, you walk away and she is waiting for you when you return, she is the one to initiate conversation with you, etc.

          But, you've put the cart before the horse. If you're already at the hand-squeezing stage, you're probably in posession of a few non-ambiguous signals. You also wouldn't use body contact to define some of the earlier stages of human interaction -- it could be completely inappropriate, and you'll seem a bigger dork. Have you tacitly been granted permission to be that close? Or are you just a bungling goon who wants to know if you touch her, she'll flinch?

          I can be socially awkward. I find it difficult to engage in social contact with new people. I can't imagine someone with a 'real' disorder being given nice codified things like you've done and be expected to apply them. Because they are, after all, subjective and hit-and-miss in terms of their predictive value.

          Even with my own 'plain old' (*) social awkwardness/geekiness, I don't think I could apply some of your cues -- at least not in the grossly simplified way you put them. There is just too much ambiguity in interpreting the responses from people, and I can't often tell where in that range something might lie.

          (*) I'm neither Autistic, nor do I have Asberger's -- but like most human behaviour, I believe it's on a continuum, and I might have some of those characteristics without actually having the affliction per se.
    • Not About Memory (Score:4, Interesting)

      by neoshroom (324937) on Thursday March 30 2006, @02:44PM (#15028808) Homepage
      It's interesting that a human could receive image data and be unable to remember what it means, but receive touch data and be able to remember its meaning.

      The issue isn't memory, its recognition. Those suffering from autism may not be able to connect to the people around them on a more emotional level, however vibration like from a ringtone is a que to stop doing whatever it is you are doing. It makes perfect sense that someone could not recognize the emotional state of another, but could easily recognize the vibration of a mobile device.
    • Re:So Simple? (Score:4, Insightful)

      by hey! (33014) on Thursday March 30 2006, @02:46PM (#15028830) Homepage Journal
      The issue is not perception or memory, but of highly specific information processing capabilities that bridge high level geometric perceptions (shapes and so forth) and awareness of the mental state of others. It's rather like having a face-recognition system that detects Osama going through airport security and rings an alarm.

      If you want to know what this is like, get married. In most cases, wives have a mechanism that alerts them to things like specks of dirt on the floor that their husbands lack. The husband can see it, it just doesn't enter his consciousness.
  • What's interesting about devices that provide behavioral feedback is that unless the user isn't aware of the device, their own actions end up reacting not only to external environmental events, but to the feedback of the device itself.

    So if the autistic user finds the device annoying, they may engage their eyes briefly to suppress the vibrating alaram. But that doesn't necessarily mean they are paying attention. Their concentration is then shifted to supressing the device.

    I am wondering if this is something
  • by robyannetta (820243) * on Thursday March 30 2006, @01:57PM (#15028278) Homepage
    "Socially challenged"? You mean WoW players?
  • by creimer (824291) on Thursday March 30 2006, @01:58PM (#15028286) Homepage
    It's about time someone tried to help the typical Slashdotter stuck in his parent's basement. :P
  • by Cr0w T. Trollbot (848674) on Thursday March 30 2006, @01:59PM (#15028297)
    For once, this really IS "news for nerds!"

    - Crow T. Trollbot

  • by PainBot (844233) on Thursday March 30 2006, @01:59PM (#15028305)
    Hopefully I can get one for my boss
  • It's my wife device.
  • by Timesprout (579035) on Thursday March 30 2006, @02:03PM (#15028361)
    So we have special key words we use so he knows when I am becoming bored or angry.

    He will say something like

    "We need to achieve synergy across our departmental endeavour so we can proactively engage any challenges the business may face"

    I will then respond

    "You are a fucking wanker"
    • > So we have special key words we use so he knows when I am becoming bored or angry.
      >
      > He will say something like
      > "We need to achieve synergy across our departmental endeavour so we can proactively engage any challenges the business may face"
      >
      > I will then respond
      > "You are a fucking wanker"

      Naaw. Just smile, nod, and shake your boss's hand.

      If the wearer seems to be failing to engage his or her listener, the software makes the hand-held computer vibrate.

      He'll get the message.

  • Nice, but... (Score:4, Informative)

    by SirBruce (679714) on Thursday March 30 2006, @02:03PM (#15028364) Homepage
    Hell, I'd like to get one of these devices for myself. But I have a feeling it would generate way too many false positives, or perhaps more importantly, inconsequential positives. The are times when people are bored, but they're trying to be polite and conversational, and pressuring them to make them more interested in what you're saying isn't going to help. Also, although this device may help an autistic person know the other person isn't engaged, do they even know what to do in order to facilitate engagement?

    Bruce
    • Let's face it. The damned thing is going to be going off all the time unless people are feigning interest.

      One of my friends does this funny thing when he's not paying attention. He breaks eye contact, starts rotating his head from side to side, scanning for something else to pay attention to. The kicker is he makes fart noises with his mouth when he does it. It's a little obvious. Funny, but obvious. I wonder if their device would pick that up.

    • See Ken Perlins page [nyu.edu] (yes, the Perlin noise guy) and check out the face applet [nyu.edu]. At the bottom there is a link to a story how it can help autistic children learn how to interpret peoples facial expressions. Best of all it's free.
  • by ma11achy (150206) on Thursday March 30 2006, @02:04PM (#15028370)

    Attaching a small camera to the side of someone's glasses isn't
    going to bode well for someone who is already socially challenged...
    • by StefanJ (88986) on Thursday March 30 2006, @02:15PM (#15028505) Homepage Journal
      Make the device look like a little anthropomorphic cricket that sits on the user's shoulder. Program it to whisper helpful hints:

      "From the way they're starting to nod off, I suspect you may have talked for a little too long about your D&D character. Maybe you should stop."

      "I could be wrong, but this guy doesn't look very interested in how parking meters are a form of statist Piracy. Maybe you should stop talking and let him finish filling out that ticket."

      "From the way she's wrinkling her nose, I suspect she thinks you smell like cat pee. Maybe you should politely back out now and think about taking a shower."
      • It could even go:

        "The person in front of you seems to be bored stiff from your rants on the evils of DRM. Would you like to..?

        • Write a letter to Sony Corp.
        • Change subject to the glory of Open Source
        • Learn more about boring people stiff

        [ ] Don't whisper in my ear again you freak"
  • "If the wearer seems to be failing to engage his or her listener, the software makes the hand-held computer vibrate."

    I thought women already had access to such vibrating devices when they got bored, although I'm not sure their use was automatic...

  • by smellsofbikes (890263) on Thursday March 30 2006, @02:07PM (#15028412) Journal
    "I see you are talking to someone who is trying to be friendly. What would you like to do now?
    * gently brush the person off?
    * actively engage the person
    * seduce the person?"
  • How long until bars start selling these in vending machines?

    When we lift the covers from our feelings, we expose our insecure spots. Trust is just as rare as devotion, forgive us our cynical thoughts. If we need too much attention (not content with being cool) we must throw ourselves wide open and start acting like a fool. If we need too much approval, then the cuts can seem too cruel. (Neil Peart)
  • by Peter Trepan (572016) on Thursday March 30 2006, @02:09PM (#15028428)
    Experts claim the "Microcomputer" will enable sufferers to hold down meaningful jobs while avoiding painful human interaction.
  • What? (Score:5, Interesting)

    by Monoliath (738369) on Thursday March 30 2006, @02:12PM (#15028456)
    I'd like to see some statistics on the accuracy of this device.

    Sounds like a horrible idea, the subject matter is so incredibly subjective, and human emotions are so incredibly fickle, laced with an infinite and exponential number of variables that determine what anything 'means' from someone, to someone else.

    Plus, does this help the autistic person learn more about people, or make them more dependent upon a machine?

    In my mind, something like this only worsens autism because it prevents the individual from having to 'learn how to understand alien stimuli' by interpreting it for them.

    I use to baby sit / care for one of my friends little brother, he was diagnosed with severe autism at an early age. Watching him grow older, in my eyes, he learned how to understand new things on his own (just sometimes it took a little longer than it does for most kids his age), like how the rest of us learn things (cause & effect / trial and error) it's not impossible for autistic individuals to perceive and comprehend this kind of stimuli, they just receive it on a different wavelength than we do, and in turn process it in a different manner.

    A device like this isn't going to 'teach' anyone anything, it's simply a crutch that IMHO, will stifle development and learning.

    As a side note, to me autism is a type of genius, that we just don't know how to comprehend as a society, this kid could do some of the most AMAZING things with number letter combinations / geometrical shapes I've ever seen.
  • Anybody remember Not Necessarily the News? Remember the one with the device that detects when you start falling asleep and beeps? As I recall, they closed the show with the then president ( Regan ) making a speech, with an audio overlay of the beeps of many journalists falling asleep.

    I think this is an amazing device ( if works as advertised, which is unlikely ) that the people who should be wearing it won't wear.
  • ...this seems to solve exactly the inverse problem to the one I see.

    In an interaction between an autistic person and a "normal" person, what I see is that it's the autistic (and/or emotionally challenged) person who gets bored and simply walks away, with little or no prior warning. I have seen this happen too many times to count. I don't think I've ever seen the "normal" person be the one to react by becoming bored; usually simply trying to comprehend what the other person is saying/doing is enough to k

  • by Chris Johnson (580) on Thursday March 30 2006, @02:24PM (#15028582) Homepage
    The question becomes, why make the user aware of this additional information?

    Expectation is that the user will apparently go, "OH! I may be boring you with my account of the history of left handed widgets! All of a sudden I don't want to finish my thought, and it mysteriously no longer matters that I haven't given you the gift of the entire intellectual structure, neatly composed with no details left out, so you can wholly share this idea that I think is the coolest thing!" ...

    I have an alternate suggestion. We should make these things so that instead of actuating a vibrating motor, the alert thing operates a small robot arm attached to a light, non damaging foam bat. When the person shows signs of boredom, the robot arm actuates the bat and whacks the listener upside the head, curing their lapse of attention and saving me the trouble. :D

    Surely this is a much more sensible approach, given that boredom is neither a virtue or considered to be a social advantage? We can teach those socially disadvantaged NTs to be socially polite even when the conversation ranges beyond 'kiddy pool' levels.

    (Disclaimer- yes, I have Aspergers, and yes, I am joking... I think ;) should I be?)
  • Wouldn't it be easier to wear a T-shirt that says:

    "If you're bored, just say so! Don't make faces at me!"
  • So basically MIT has developed.. um.. Booze

    Oh wait, that would be a mesication, would the device be shot glasses?
  • by Odocoileus (802272) on Thursday March 30 2006, @02:29PM (#15028633)
    It might be nice to know when I am losing someone's interest, but, as an Aspie, I really don't have much to say to NT's anyway. I mean if I could hold conversations that interest a NT I wouldn't need the device in the first place. The reality, however, is that conversations that seem to intrigue NT's hold no interest for me. And for some reason I do not get, NT's do not like to talk about the same couple of topics incessantly. I have learned to do the obligatory greetings, but they are best kept short. Anything else is either about business, which has a finite set of interactions (I am fine within my knowledge base), or involves friends that have similar interests. I know some aspies want better communications with the NT world, but knowing when the person is bored would, at least for me, be worse because I still wouldn't know what words to speak to make it better. I guess in the long run maybe, after performing some statistical analysis concerning what words make a person bored. But then again, I pretty much already know that people do not want to talk about scifi or computers or world domination, so it is back to square one.
  • Wouldn't it be easier for the autistic person (given that they actually can engange and talk to people so much that those could become bored) to simply train how to recoginze different facial expressions? Don't flame me please, I know it ain't easy and maybe for some it can't be done. I simply think if an autistic person that is able to engage people on their own might want to try and learn facial expressions using different means. Like we don't naturally know how a dog feels. But if we learn that tail wagg
    • You're really taking your diagnosis of paranoid schizophenia a bit harshly, aren't you?

      NOTE: I'm not part of the conspiracy against you. I promise.