Anyone who has been surfing the web for a while is probably used to clicking through a CAPTCHA grid of street images, identifying everyday objects to prove that they’re a human and not an automated bot. Now, though, new research claims that locally run bots using specially trained image-recognition models can match human-level performance in this style of CAPTCHA, achieving a 100 percent success rate despite being decidedly not human.

ETH Zurich PhD student Andreas Plesner and his colleagues’ new research, available as a pre-print paper, focuses on Google’s ReCAPTCHA v2, which challenges users to identify which street images in a grid contain items like bicycles, crosswalks, mountains, stairs, or traffic lights. Google began phasing that system out years ago in favor of an “invisible” reCAPTCHA v3 that analyzes user interactions rather than offering an explicit challenge.

Despite this, the older reCAPTCHA v2 is still used by millions of websites. And even sites that use the updated reCAPTCHA v3 will sometimes use reCAPTCHA v2 as a fallback when the updated system gives a user a low “human” confidence rating.

  • mynamesnotrick@lemmy.zip
    link
    fedilink
    English
    arrow-up
    97
    arrow-down
    1
    ·
    1 month ago

    Need these bots as a browser addon now. When your using a VPN these things are the bane of your Internet browsing experience.

  • Fester@lemm.ee
    link
    fedilink
    English
    arrow-up
    52
    ·
    1 month ago

    So the robots are now more successful at proving they’re human than I am.

    • OpenStars@discuss.online
      link
      fedilink
      English
      arrow-up
      11
      ·
      1 month ago

      The inverted Turing test, it would seem.

      Beep boop I am smarter than you… 🎵

      Bleep bloop you seem to me dumb as rock… 🎼

      img

  • zephorah@lemm.ee
    link
    fedilink
    English
    arrow-up
    31
    ·
    1 month ago

    I need a bot then. I click through 7-8 of those and still don’t pass.

  • merc@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    31
    arrow-down
    1
    ·
    1 month ago

    What’s ironic is that the main purpose of reCAPTCHA v2 is to train ML models. That’s why they show you blurry images of things you might see in traffic.

    AFAIK the way it works is that of the 9 images, something like 6 are images the system knows are True or False, and another 3 are ones it is being trained on. So, it shows you 9 images and says “tell me which images contain a motorcycle”. It uses the 6 it knows to determine whether or not to let you pass, and then uses your choices on the other 3 to train an ML model.

    Because of this, it takes me forever to get past reCAPTCHA v2, because I think it’s my duty to mistrain it as much as possible.

    • Kuvwert@lemm.ee
      link
      fedilink
      English
      arrow-up
      16
      ·
      1 month ago

      You’re wasting your time. Your fingerprint is graded and discarded if you’re not reliable

      • merc@sh.itjust.works
        link
        fedilink
        English
        arrow-up
        3
        ·
        1 month ago

        At least it adds noise to the system. It’s better than the people who are happily training the AI.

  • dan1101@lemm.ee
    link
    fedilink
    English
    arrow-up
    22
    arrow-down
    1
    ·
    1 month ago

    That’s good, hopefully we will quit seeing them. Because I’ve gotten to the point if I see a captcha I just go to another site unless it’s something I’ve got to do.

  • ikidd@lemmy.world
    link
    fedilink
    English
    arrow-up
    13
    ·
    1 month ago

    Then give me a FF addon that does these for me because I’m nowhere near that successful.

    • ChaoticNeutralCzech
      link
      fedilink
      English
      arrow-up
      5
      ·
      1 month ago

      Google has. The Zürich university fella needed his own data classification monkeys.

  • werefreeatlast@lemmy.world
    link
    fedilink
    English
    arrow-up
    7
    ·
    1 month ago

    Recaptcha 4.0… what do you think about this image…an image of a kid riding their bike without any protective gear on a freeway.

    AI: a bike with a kid on it on a road. Perfectly fine.

      • T156@lemmy.world
        link
        fedilink
        English
        arrow-up
        3
        ·
        1 month ago

        It’d be a bit unreliable, though. Not everyone has the same reaction to the same thing, nor do they express it in a similar way.

        Someone might think a snake or a spider is cute, whereas another would want to incinerate it on the spot. A third might be concerned because they seem to be injured, etc.

        Not to mention that image recognition/emotional analysis has been an ongoing field of research for some time. Making the link is not overly difficult.

  • Maggoty@lemmy.world
    link
    fedilink
    English
    arrow-up
    6
    ·
    1 month ago

    Oh good we finally trained them enough. Can we please get rid of captcha now have the next portion?