this is also can be done with tesseract or encog framework... i dont know whats news about this
Google's Street View computer vision can beat reCAPTCHA with 99% accuracy
101–110 of 159 posts
Re: Google's Street View computer vision can beat reCAPTCHA with 99% accuracy
#102Earlier quoted context omitted.
It has been like this for a while now. AFAIK you have either - the actual captcha and a word from a book (Google Books) - the actual captcha and a house number from Street View (Google Maps) Only the actual captcha has to be typed in correctly.
Do people ever intentionally type in nonsense for the non-captcha part, out of spite, for instance? If you could persuade enough people to provide non-answers for the non-captcha part, would Google stop using it (the non-captcha part)? My guess is it's like spam. As long as it works a small fraction of the time, it will still exist.
At one time, there was this effort on 4chan to spam recaptcha with "penis" for the second word. When this got to some online news site, they asked the recaptcha guys who confirmed that the traffic was irrelevant.
I suspect the nonsense won't matter. You're drowned out.
Re: Google's Street View computer vision can beat reCAPTCHA with 99% accuracy
#103In effect, Google computer vision got so good that they made their own system obsolete. This is a good thing. I still think the only reliable way to confirm identity (or humanity) online is an email or SMS verification. Recently, receiving a 2-factor SMS code took less time than the page refresh prompting me to enter it.
Wouldn't it be trivial to code up a Twilio-powered bot to defeat the SMS verification?
Re: Google's Street View computer vision can beat reCAPTCHA with 99% accuracy
#104This particular issue (AI performance on captchas) is really quite fascinating. It's an arms race, but the problem is, only one side can win. Google is claiming they have improved their system in some (understandably) unspecified way, but there's only so far this can go. Captchas need to detect whether someone is human, but it has to work for everyone, ideally, even those with disabilities. Any simple task a human ca…
> It's an arms race, but the problem is, only one side can win. I don't disagree, but although this specific problem can be automated, the arms race will continue as newer problems (requiring uniquely human intelligence) are presented. We can keep moving the bar until the singularity. For example, a captcha: "Pick the most unsafe environment", followed by pictures of a house, a park, a volcano, a bed. To solve this,…
The odd thing with a question like that is how "most dangerous" is determined. Logically the volcano is the most unsafe environment, and people do die on them, but statistically far more dangerous, at least by death count, is the bed.
This is the kind of problem that would still cause issues for some people, especially if you had to scale them to produce identical captcha's relatively rarely (if the same one shows up too many times it can easily be hard-coded, requiring the most basic detection to solve it reliably) in that as you come up with more of these types of questions the expected answer gets rapidly more nuanced.
Re: Google's Street View computer vision can beat reCAPTCHA with 99% accuracy
#105In effect, Google computer vision got so good that they made their own system obsolete. This is a good thing. I still think the only reliable way to confirm identity (or humanity) online is an email or SMS verification. Recently, receiving a 2-factor SMS code took less time than the page refresh prompting me to enter it.
Requiring SMS or email creates problems for users that wish to remain anonymous. SMS is essentially not an option and the number of free email providers that accept registrations sans-SMS via tor continues to shrink.
Re: Google's Street View computer vision can beat reCAPTCHA with 99% accuracy
#106Earlier quoted context omitted.
> It's an arms race, but the problem is, only one side can win. I don't disagree, but although this specific problem can be automated, the arms race will continue as newer problems (requiring uniquely human intelligence) are presented. We can keep moving the bar until the singularity. For example, a captcha: "Pick the most unsafe environment", followed by pictures of a house, a park, a volcano, a bed. To solve this,…
The odd thing with a question like that is how "most dangerous" is determined. Logically the volcano is the most unsafe environment, and people do die on them, but statistically far more dangerous, at least by death count, is the bed. This is the kind of problem that would still cause issues for some people, especially if you had to scale them to produce identical captcha's relatively rarely (if the same one shows up…
So it's a captcha that keeps out computers and pedants on HN. Win/win, in my opinion.
Re: Google's Street View computer vision can beat reCAPTCHA with 99% accuracy
#107This particular issue (AI performance on captchas) is really quite fascinating. It's an arms race, but the problem is, only one side can win. Google is claiming they have improved their system in some (understandably) unspecified way, but there's only so far this can go. Captchas need to detect whether someone is human, but it has to work for everyone, ideally, even those with disabilities. Any simple task a human ca…
> It's an arms race, but the problem is, only one side can win.
It's already happening.
It turns out humans come quite cheaply, and you can actually solve captchas using humans at very high QPS for very low cost. And people do!
Site owners will have to come up with alternate means of "bot" enforcement that does not rely on human vs computer detection.
Re: Google's Street View computer vision can beat reCAPTCHA with 99% accuracy
#108This particular issue (AI performance on captchas) is really quite fascinating. It's an arms race, but the problem is, only one side can win. Google is claiming they have improved their system in some (understandably) unspecified way, but there's only so far this can go. Captchas need to detect whether someone is human, but it has to work for everyone, ideally, even those with disabilities. Any simple task a human ca…
> It's an arms race, but the problem is, only one side can win. I don't disagree, but although this specific problem can be automated, the arms race will continue as newer problems (requiring uniquely human intelligence) are presented. We can keep moving the bar until the singularity. For example, a captcha: "Pick the most unsafe environment", followed by pictures of a house, a park, a volcano, a bed. To solve this,…
A bot would solve this one with a random selection 25% of the time. You would need multiple puzzles to make this a useful solution.
Re: Google's Street View computer vision can beat reCAPTCHA with 99% accuracy
#109Earlier quoted context omitted.
> It's an arms race, but the problem is, only one side can win. I don't disagree, but although this specific problem can be automated, the arms race will continue as newer problems (requiring uniquely human intelligence) are presented. We can keep moving the bar until the singularity. For example, a captcha: "Pick the most unsafe environment", followed by pictures of a house, a park, a volcano, a bed. To solve this,…
"Pick the most unsafe environment", followed by pictures of a house, a park, a volcano, a bed. My sense is that IBM's Watson could tackle this particular question. Also interesting to note that the question introduces external considerations, such as whether or not there's someone dangerous near the house, bed, park, etc. A human respondent will have to ignore the matter of whether there's additional context and seco…
Of course the simple way around
"Pick the most unsafe environment", followed by pictures of a house, a park, a volcano, a bed.
if you're a spammer is to choose randomly and make four times as many attempts. Which is why catchpas make you type six or more letters giving millions of combinations so you can't do it randomly. To get up to a similar resistance to random attempts with 'pick the unsafe environment' type questions you'd need of the order or eight of them (giving 4^8 or approx 65k combinations). Would you as a user want to fill in 8 of those things?
Re: Google's Street View computer vision can beat reCAPTCHA with 99% accuracy
#110This particular issue (AI performance on captchas) is really quite fascinating. It's an arms race, but the problem is, only one side can win. Google is claiming they have improved their system in some (understandably) unspecified way, but there's only so far this can go. Captchas need to detect whether someone is human, but it has to work for everyone, ideally, even those with disabilities. Any simple task a human ca…
I wonder about the computer's ability to figure out what's spam and to manage the visibility of posts from new accounts dependent on its confidence in that. If captcha proves unwinnable, which I agree with you it very well may be, then all isn't necessarily lost.