Hi! Please, guys, take a look here: http://gestcha.com - this is a hand gesture CAPTCHA prototype, that I finished just yesterday (it works). Please, tell me, could it be useful or just a waste of time?
Now that’s more secure than captcha, great stuff. Just that would drive away few people. Captcha is simpler to understand, you don`t have to read the description and waste time by doing so. Nevertheless, i like the idea! Connect corresponding dots.
I like the idea, but [LIST]
[]it requires use of a mouse, and is therefore not suitable for keyboard-only users
[]an alternative would also be required for visually-impaired users
[*]it requires good motor/co-ordination skills. (I failed on one attempt for not placing the dots accurately enough.)[/LIST]
I failed every time and I think the problem was I could not tell on the photo which finger was which as the photos were very dark. It might have been the problem TecnoBear had in the spot placing accuracy.
It looks like a good idea and worth perusing. I would suggest snap points on the fingers when you get close - not necesselary snap points for just the correct answer but even if the answer is wrong and better quality photos. Although this then makes it simpler for the robots?
Hmmmmm…I failed every time, and it sure looked like I hit the right fingers every time. Is it limited to specific points on the fingers perhaps? Is the target area too small for passing the test?
Better quality and larger photos might be the answer to making it easier for humans but harder for robots. Humans would benefit from having a larger and clearer target area, while bots would be hampered by a larger “wrong” area.
Hi! Thank you, guys, for your attention! OK, I need to explain some things: Actually an answer matrix is 4x4. Please, wait 10 min. - I will activate the grid. Thank you, TechnoBear! You gave me the great idea about keyboard-only devices (TVs, etc.). It’s not hard to add the numbers (1,2,3…) to cells and check the buttons. Thank you, once more! About touch-fingertip proximity: actually, I can do it less strict - it will put some extra weight on generation process, but I can afford it. But it will borrow some time. Thank you, everybody! It was a right place to visit first!
P.S. Grid is activated. You may need to update page in browser (I don’t know what I’m doing wrong with it - I have to press “update” in Firefox to see an updated page). Proximity to edges can happen be too small sometimes - you were right. Not a problem. But keyboard support needs some more time to add.
That’s better; I failed three out of three the first time and one out of three this time
Concerning brightness - I can make it lighter, but with detriment to dynamic range. Is it really worth it? I have doubts. Size is unchangeable - sorry. Actually, it was chosen to be 96x96 due to many factors (hardware-related also).
To Rubble: Give me one hour, please - it will be 10 of 10 
It took longer, but it’s done - should be much better now.
Personally, I think it’s too hard. I failed the first attempt, and that would be it for me if the wild. I’m so fed up with capthcas and being treated like a miscreant that I tend to move on when treated like this. It’s not clear which finger is which, and those with seeing/cognitive difficulties wold have no chance to get through this barrier.
I still feel that any captcha that treats the honest person as the criminal is not the way to go. It’s better to have it going on in the background, like a minimum time for filling out a form or whatever.
You’ve left me no chance for justification :). It’s a little bit offtopic, but I’m so agree with you! You’re right - as it usually happens: to get rid of something bad, people invent something worse :). But concerning the last suggestion - IMHO, it won’t work reliable enough (and to my deep regret). Remember just recent event - “Eugene Gustman”. It just talks for itself. Moreover, I believe, that my graphic CAPTCHA will also be broken one day, and It will be the great moment, by the way: gesture recognition is a challenge for AI nowadays. But, enough of that, please, give me the real hint, what should I do to make my disgusting CAPTCHA even better!
Heh, my only hint, as I said, is to hide it from legitimate users. These days, I usually don’t bother to complete a form when I see a captcha at the bottom, as I find them too difficult/buggy, meaning that it’s a waste of time filling out the form.
No objections. I’m agree, there is too much of it everywhere, even when it’s not so necessary.
I do think that the idea is innovative and creative, and it’s certinaly some new. Who knows if it will become a standard?
I only tried it once but although it took me a few minutes to work it out, I did fine (I passed!). It is true that it was difficult to recognized each finger but with a bit of patience, I did it.
The pictures are so dark that it makes it more difficult. So, yes, there’s room for improvement and I’m not telling you anything that you don’t know yet. Just confirming what you know.
The next generation could be a benefit for phone users. Just put your finger and I’ll take your fingerprint… though in a few years this may be a way to control who visits which.
Thank you, guys. It wasn’t planned, but it happened to be a sort of beta-testing :). To be honest, the main factor in the development process was to make it as hard for all kind of future bots as possible. Today, I clearly understand the excess of the most of measures for the present moment, so, I give them up one by one (or postpone, better to say). Sure, there is also some space for higher brightness (light sources can be adjusted in certain way) and so on. So, thank you once again! Please, visit my site from time to time - I’ll be trying to improve the engine constantly. Welcome!
Some news: 1. Just made big and important modification: completely different check-subsystem now, so no matrix is used there - answer accuracy is now controlled by endpoint site mechanics only and can be set as presize as 1px! (could be suitable for real big screens only, equipped by high-dimensional pointing devices, though brute-force attack seems to be practically useless there). It’s 8px now - seems optimal. 2. Brightness was increased (much better for energy-saving display configurations, but I feel scenes lack some depth now). 3. Keep thinking about some comfortable input method for keyboard-only devices - nothing worthwhile yet.
Just to make sure this is clear to everyone: it is not a test for humanness. I haven’t found a CAPTCHA yet that tests if someone is a human. And, humans send spam.
I won’t do any monkey dances (captchas) for any sites that are not absolutely necessary, like getting access to government services or public transport. I’ve skipped joining a few forums that looked interesting because they had CAPTCHAs, which is a group of letters that translates to
“We have a spam problem, and we require our users to fix it for us, because they are lowly users and we would rather make them do a monkey dance than we actually do all the well-known techniques we ourselves are responsible for doing instead.”
I did look at the page, but couldn’t figure out what exactly I was supposed to do. I’m probably too stupid for such a thing, which prolly means this works perfectly: you wouldn’t want stupid people like me commenting on your whatever website.
Hi! What can I suggest: just try to treat it as a game! It was a main goal for me here - turn the pain from recognizing of that damn meaningless mess of crippled symbols into something much simpler and even fun may be. Moreover, recent experience tells me, that noticeable learning effect is taking part! For regular captchas I found no such thing (may be I’m stupid). So, take it easy and try!