Rumor: Apple intends to replace Touch ID fingerprint scanner with advanced facial recognit...

1235»

Comments

  • Reply 81 of 91
    StrangeDaysStrangeDays Posts: 12,877member
    So how does one use facial recognition in retail stores for Apple Pay? 
    Push the virtual home button and let the sensors do their thing on your face, from an angle, as the rumors have suggested it will be able to do? Don’t seem the problem. 
    So they’re going to be able to read my face while I have my phone pointed at the POS. In some stores where I use Apple Pay I know that wouldn’t work.
    I’ve never used a POST that wasn’t at approximately table top level, which the Gurman rumor said it can do. Much like the document scanning feature in iOS 11 beta today. 
  • Reply 82 of 91
    Rayz2016Rayz2016 Posts: 6,957member
    Some fella hit the nail on the head a few months back. He said something like:

    Your face is the user name. 
    Your thumb print is the password. 

    And he was absolutely right. I think this is about bringing multi-user capabilities to iOS. 
    avon b7patchythepirateanton zuykov
  • Reply 83 of 91

    They could do a two camera solution that would sense depth to get around the Xerox thing. Though in this case if anyone wants to unlock your phone they just have told hold it to your face. I kind of wonder if this is going to enable a lot more law enforcement types to get around bugging Apple to unlock a phone simply by having the person have to stand in front of it.

    This latter question is why I will disable any such feature for unlocking my phone.  Like fingerprints, the courts will decided that one's face isn't "protected" by the Fifth Amendment the way a password is.  They can force me to stand in front of it or touch it all they want, my phone won't unlock.
  • Reply 84 of 91
    melgrossmelgross Posts: 33,510member
    Soli said:
    zoetmb said:
    IMO, an incredibly silly conversation.  Even if there is a very slight chance that two human beings have the same fingerprints (or close enough for the iPhone to make an error), what is the chance that the wrong one is going to get access to your iPhone?    The chance that anyone is going to get hold of my iPhone is incredibly small unless I'm held up at gunpoint. 
    This has been answered. According to Apple, it's 1:50,000.

    But addressing what you call "silly," since your fingerprints are all over the device, likely with your Touch ID fingerprint on the sapphire of the Home Button, it's not, as you say, "silly" to make the parameters of authentication more stringent so that a false-positive is less likely to occur. Also keep in mind :that you get far fewer tries with Touch ID than you do with a PIN, and that you can't use Touch ID without having at least a PIN, without inputting that PIN before Touch ID is active after a restart, and re-inputting your PIN once Touch ID has failed a handful of times in a row. This "silliness" of approximation v precision is the reason the a memorized code has priority over a biometric.
    Did you even bother to read my post to you? It’s not 50 thousand, it’s 25 trillion, or more. I don’t know where you, or Apple, came up with that number, but it’s completely wrong, even for the Rez sensor they’re using.
    edited July 2017
  • Reply 85 of 91
    igorskyigorsky Posts: 755member
    rob53 said:
    I can't see facial recognition working consistently for very many people. I understand how it works but people wear sunglasses and other things that could affect the geometry of the face. I also can't see people holding their phone up to their face and hoping everything works. I thought Samsung's attempt at facial recognition was broken by a simple xerox of the users face. I hope Apple takes their time before they change so they make sure it works all the time and isn't susceptible to even advanced attempts at faking a face to get access. I know people say they're faked fingerprints but they're not as easy as it sounds.
    First of all it's hilarious that you feel Apple will release a facial recognition system with the same issues that Samsung is having; these two companies are on different levels when it comes to engineering. Second, I've read that Apple's 3D facial recognition system will use more data points than Touch ID, thus making it even more secure.

    I really don't understand the panic I'm reading on the internet, if this rumor is even in fact true. Security is Apple's bread and butter...I have zero doubt that they'll get it right.
    edited July 2017
  • Reply 86 of 91
    melgrossmelgross Posts: 33,510member

    They could do a two camera solution that would sense depth to get around the Xerox thing. Though in this case if anyone wants to unlock your phone they just have told hold it to your face. I kind of wonder if this is going to enable a lot more law enforcement types to get around bugging Apple to unlock a phone simply by having the person have to stand in front of it.

    This latter question is why I will disable any such feature for unlocking my phone.  Like fingerprints, the courts will decided that one's face isn't "protected" by the Fifth Amendment the way a password is.  They can force me to stand in front of it or touch it all they want, my phone won't unlock.

    I don’t get why some people are so worried about the courts. Are you doing something illegal? Are you planning to do something illegal? If so, then I want you to get caught. If you’re not, then what’s your problem?
    gatorguy
  • Reply 87 of 91
    anton zuykovanton zuykov Posts: 1,056member
    Apple will not "replace" Touch ID with facial recognition.
    That is just virtually not possible, due to it being prone to problems, the touch id lacks.
    At the worst case, they might ADD it to the existing touch id tech, but replacing it would just be like shooting your own arm, because sleeves of your shirt need holes for ventilation.
    I mean, come on, how would you use facial recognition in the darkness? IR diodes? Ok, fine, but you would have to turn your phone toward your face precisely, so that the camera array can "see" your face. That is real convenient! What happens when your phone has happened to register someone else's face (by accident). Would it consider it to be false attempt or would it just ignore it?
    There is just too many ways this could go wrong, while offering no discernible benefit over the existing touch ID auth tech.
    You know shit about what Apple has been working on. 

    Prepare to be wrong.
           You don't need to know exactly that a company A is working on, to predict which supposed changes will NOT happen. For example, we know that the next iPhone will not have a portable fusion reactor, instead of a battery. Why? Because we know that that tech is NOT THERE yet (not even close).
    The same logic applies here - face recog tech has A LOT of problems. And some of those problems come from the fact that scan of a face has to be obtained.
    radarthekat
  • Reply 88 of 91
    anton zuykovanton zuykov Posts: 1,056member
    jahaja said:
    You guys see only problems. Here are some possible solutions, without ever having studied the subject (so probably there are 100 times better ones, if you do work on it for years):
    1) Double cameras can sense depth, thus no xerox can fool it.
    2) Heat sensor can add further security versus a piece of paper.
    3) Digital compensation for perspective means your face wouldn't have to be completely parallell to the phone surface for recognition.
    4) IR camera could compensate for darkness.

    In the European Union, facial dimensions are since long included digitally in passports – fingerprints are NOT.
    1) correct
    2) Can easily be circumvented by heating a fake face, that you are trying to "recognize" with that tech.
    3) For 2D surface that technique works fine. However, it will not work, because you just proposed to use a 3D surface. If certain features are obscured (due to a face being turned a bit sideways, they are, well, obscured and no amount of further transformation will restore that missing data.
    radarthekat
  • Reply 89 of 91
    SoliSoli Posts: 10,035member
    melgross said:
    Soli said:
    zoetmb said:
    IMO, an incredibly silly conversation.  Even if there is a very slight chance that two human beings have the same fingerprints (or close enough for the iPhone to make an error), what is the chance that the wrong one is going to get access to your iPhone?    The chance that anyone is going to get hold of my iPhone is incredibly small unless I'm held up at gunpoint. 
    This has been answered. According to Apple, it's 1:50,000.

    But addressing what you call "silly," since your fingerprints are all over the device, likely with your Touch ID fingerprint on the sapphire of the Home Button, it's not, as you say, "silly" to make the parameters of authentication more stringent so that a false-positive is less likely to occur. Also keep in mind :that you get far fewer tries with Touch ID than you do with a PIN, and that you can't use Touch ID without having at least a PIN, without inputting that PIN before Touch ID is active after a restart, and re-inputting your PIN once Touch ID has failed a handful of times in a row. This "silliness" of approximation v precision is the reason the a memorized code has priority over a biometric.
    Did you even bother to read my post to you? It’s not 50 thousand, it’s 25 trillion, or more. I don’t know where you, or Apple, came up with that number, but it’s completely wrong, even for the Rez sensor they’re using.
    I did. You wrote, "In my quick search, I did find a number that said about 25 trillion. But that number is defined by DNA combinations of the relevant factors."

    I think I'm going to go with Apple over your speculative and unsourced comment about 25 trillion, especially when it references "DNA combinations," as well as talking about PPI of the sensor without any reference to precision of analysis or number minutiae, which can be  adjusted as Apple desires for their security-through-convenience feature. Even DNA, which is very specific chains that are not unlike a very long, BASE-4 passcode, still don't get results that are 100% certain. Example: "99% sure he's not the father."

    If you really think that Touch ID is that much more secure than a passcode, then why do you believe that everything defaults to the passcode (i.e.: the more secure element in the chain)?
  • Reply 90 of 91
    melgrossmelgross Posts: 33,510member
    Soli said:
    melgross said:
    Soli said:
    zoetmb said:
    IMO, an incredibly silly conversation.  Even if there is a very slight chance that two human beings have the same fingerprints (or close enough for the iPhone to make an error), what is the chance that the wrong one is going to get access to your iPhone?    The chance that anyone is going to get hold of my iPhone is incredibly small unless I'm held up at gunpoint. 
    This has been answered. According to Apple, it's 1:50,000.

    But addressing what you call "silly," since your fingerprints are all over the device, likely with your Touch ID fingerprint on the sapphire of the Home Button, it's not, as you say, "silly" to make the parameters of authentication more stringent so that a false-positive is less likely to occur. Also keep in mind :that you get far fewer tries with Touch ID than you do with a PIN, and that you can't use Touch ID without having at least a PIN, without inputting that PIN before Touch ID is active after a restart, and re-inputting your PIN once Touch ID has failed a handful of times in a row. This "silliness" of approximation v precision is the reason the a memorized code has priority over a biometric.
    Did you even bother to read my post to you? It’s not 50 thousand, it’s 25 trillion, or more. I don’t know where you, or Apple, came up with that number, but it’s completely wrong, even for the Rez sensor they’re using.
    I did. You wrote, "In my quick search, I did find a number that said about 25 trillion. But that number is defined by DNA combinations of the relevant factors."

    I think I'm going to go with Apple over your speculative and unsourced comment about 25 trillion, especially when it references "DNA combinations," as well as talking about PPI of the sensor without any reference to precision of analysis or number minutiae, which can be  adjusted as Apple desires for their security-through-convenience feature. Even DNA, which is very specific chains that are not unlike a very long, BASE-4 passcode, still don't get results that are 100% certain. Example: "99% sure he's not the father."

    If you really think that Touch ID is that much more secure than a passcode, then why do you believe that everything defaults to the passcode (i.e.: the more secure element in the chain)?
    Those were two different sources. The 25 trillion was a scientific formulation, and there’s no reason to disbelieve it. The second was, if you read it (which I can see that you didn’t), a very well researched, and illustrated article from the university. You’re disregarding information which is far better than the “Apple says it’s 50,000.” Which is a number that no research on the subject would agree with.

    The reason why it defaults to a passcode:

    http://www.macworld.com/article/3072181/ios/new-touch-id-rules-why-you-have-to-enter-your-passcode-when-you-wake-up.html

Sign In or Register to comment.