Face ID

This topic contains 19 replies, has 2 voices, and was last updated by  Mush 1 week, 5 days ago.

Viewing 15 posts - 1 through 15 (of 20 total)
  • Author
    Posts
  • #31073 Reply

    Mush

    Is this used for Face ID? Someone told me a similar technology is used for it

    #31142 Reply

    Ray

    What?

    The mechanisms used in the library are most commonly used in random generation based on a seed. The most obvious example is Minecraft.

    Face ID is most likely going to be a machine learning technology that builds up facial recognition.

    #31151 Reply

    Mush

    But we can permute the combinations of faces into recognizable algorithms derived from the mechanical generator of the image archives… right?

    #31161 Reply

    Rakesh Maharjan

    How can I get my future

    #31162 Reply

    Rakesh Maharjan

    Black and white

    #31167 Reply

    Ray

    No, not really. You’d need to generate faces first, and then find where they are located in the image archives. You might as well just generate faces, the image archives have no purpose here.

    What do you think the image archives are doing that would be specific to faces?

    #31422 Reply

    Mush

    the image archive can be searched, yes? but it has all images already. so we can look at faces and there would be matches to those faces in the archive. right? so why can’t we use it for face id.

    #31437 Reply

    Ray

    Because for every match of a face, there are a million matches of slightly off and nonexistent faces. The search function will also match literally any image you give it, not just a face. Because the image archives can find where a face is located, that doesn’t give you any new info. It just knows that this particularly image of a face exists in that location.

    To make this simpler, do you understand why the text section of the library provides no knowledge?

    #31438 Reply

    Mush

    lol text has all the answers ever… if you know where to look. which you do, because of your numbers formula thing which you said. the text section has all knowledge ever. to say it contains no knowledge is a lie. the knowledge that you might extract out of it might be nothing, but that’s only because you don’t know where to look.

    #31441 Reply

    Delengroth
    Participant

    The thing about face ID is that, whatever program you’re using to match with, it will already have its own database of faces associated with names. If you want to use the Library to find alternate angles or photos of a person that don’t (naturally) exist of a certain person, then I suppose that would be an interesting to utilize it.

    The other problem though, is that there are a mind-bogglingly huge amount of images to sift through. It can take trillions of years just to maybe even find anything noteworthy, and even then you’ll have to compete with every possible photograph that can be taken. Imagine if after a trillion years, instead of finding a face, it finds a distorted and out of focus picture of someone’s dinner plate that they wanted to post on Instagram.

    #31486 Reply

    Ray

    That fact that you need to know where to look indicates that no knowledge is provided. If the cure for cancer can be made using specific instructions, I can’t actually find those instructions without already knowing them. Hence, I don’t learn anything new if I find something in the library.

    If the library were truly to provide knowledge, you would need to find information without already having that information.

    But I need you to be more specific with face ID. What would the image archives do for face ID that can’t be done with a database and machine learning? At what step of the process do we include the image archives?

    #31493 Reply

    Mush

    we put the image library where we search the face up in the is gov facial database.

    #31519 Reply

    Ray

    Ok, that might be useful, but most images are already stored on websites, removing a need to archive them like that. Facial recognition also has a lot less to do with the actual pictures of someone’s face, and a lot more to do with the things AI can create by looking at pictures of faces. Thus, storing the picture of a face is pretty unimportant if the AI has already seen it.

    #31809 Reply

    Caz

    If you’re worried about facial recognition software, do something like cover part of your face with your hand, or hair. The software itself identifies people using face shape and general curvatures/protrusions etc. It’s why certain asymetrical haircuts can confuse most recognition software.

    #31810 Reply

    Mush

    I am a proprietor of facial software by copious,overflowing amounts in my database and I want to use Monsieur Basile’s great deal to help extend the capabilities of my regimentary formation. I don’t need to cover my hand with my hair, it’s much better to enter my whole face into the database and i need the image archives to do so. ca va?

Viewing 15 posts - 1 through 15 (of 20 total)
Reply To: Face ID
Your information: