Talk:CAPTCHA

Other related discussions triggered by GSoC 2014: [1] [2].

Possible solution to Multilingual, usable and effective captchas

While researching i found that Wiki Commons could be act as a database to wide no of images but since there are non tags associated with the images and categories the image opt for is highly unreliable (for example in the category "cat" contains several vague images which does not have cat in it instead they are mentioned in the category cat while retrieving images these vague images can also act as correct answer)....so it could create problem for images they have options as. For the solution i figured out three ways to encounter the problem: 1)Use image recognition:To determine the whether a particular image contain object it is looking for (example:if the captcha asks a question to select the cat; the image recognition tool will then determine whether the image given as option contain cat).This image recognition tool can be build with the help of Python with opencv library since Php's GD library is not ideal for image recognition also it is much slower. 2)Use Clip Arts:We can make a database of clip arts of various objects and making a combination of them and asking the questions accordingly for example:we can use clip arts of monkey,cat and tiger make their combinations and ask the user to select the image with tiger and monkey. 3)we can use first five images of a definite category as an options but we will be left with very few options in our hand

In my opinion image recognition is the best possible solution to the problem since it would also help us with annotations as mentioned here and advantage to image recognition can also be that it could act as an important pillar for Wiki commons as i personally fell that its database is not well organized

Your suggestions are highly recommended and welcome. Cheers ~aalekhN AalekhN (talk) 13:38, 25 February 2014 (UTC)

Captchas need to be hard to solve by machines. If machine recognition can be used to infer the categories when creating the captcha, it may be also used to solve it.

The captcha system does not need to pick from all possible images on Commons, it is fine to use just a subset of those that works, and even make the system to improve over time (discarding the images for which users had asked to reload the captcha or users fail to resolve). Pginer (talk) 14:56, 25 February 2014 (UTC)

Captcha: a newer idea

hello,

These are some approaches i can think of instead of a text based captcha.

The image idea where users are asked to spot the odd one out like demonstrated or find all the similar images like mentioned in here.

Also a picture with a part chipped in could be shown and chipped pictures could be given as options

like find the missing part from a jigsaw puzzle.

The image which would be shown is http://imgur.com/uefeb08

http://imgur.com/KEJqCg3 is the picture which would be the correct option.

The other options could be rotated versions of this , which would not be so easy for the bot to match. (unless it somehow worked some digital processing algorithm and matched the color gradients or something like that).

This is a good option for people who do not know english or are illiterate and maybe would not understand questions like : is this a bird , plane , superman? after being shown a picture.

Tell me what you think

(Sorry to upload those images on imgur. i dont know how to put them on the wiki .Hope that is ok) Gmansi (talk) 12:08, 27 February 2014 (UTC)

This got a reply on wikitech-l by Anomie, such puzzles are easily solved by bots he says. Nemo 22:11, 6 March 2014 (UTC)

New Captcha Idea

1)Alphabetical order captcha:We can use Html5's drag and drop Api to list a particular Set of images into one category .for example in the example mentoined in the demo here ,i made a collection of diff words starting with letters A,B,C ....as an output i grouped up words with starting letter A diff from words with starting letter B,C....As,i used text in this example we can use images of diff animals such as cat's and dog.....and by drag and drop we can group images of cat and that of dog in diff categories.

2)Annotation captcha:We can use Images With annotations from commons determine the subcategoriy the annotations belongs to and then give relevant options to the users ;for example in the file we can search from names of different annotations to which they corresponds to from wikipedia(names given here are those of mountain) and then give the the option's much relevant to the image.

3)Effect captcha:We can use image as a question which are changed by the effect produced php's gd library and the use the same file with another effect and then ask user to match the two files.....for example:the image1 can be used as a question asking user to click on the image that matches with the question image and as an answer we can give this spiral image of the original image.

Similarly we can give filters to different images producing different options asking user for right answer.

4)Direct captcha:We can ask to user direct questions like ask for selecting cat out of options consisting of images of cats and humans.....an example by pginer demonstrate this example

5)Ask User to click on given effect: Asking user to click on images consisting of spiral effect's out of options which consist of images with spiral and other effects(example:greyscale).

6)Drag and Drop character in Correct Place: We can use drag and drop api of html5 to ask user to form an particular alphabet or no out of the pieces of character provided .Here is an example to form an character "A" and an digit "8" out of the same pieces of character.

This drag and drop capability can be further enhance to form a particular shapes.For example form a clip art from a particular set of piece of shapes....for example the image given here inserts the correct nose as asked in the in the questions out of the possible options provided.

Most,Importantly i think creation of an index system would be fruitful since it would rank the inappropriate images on the basis of users response (rank is negative for an image if user needs to reload a captcha) to a provided captcha.This as the time passes will provide us with relevant images which are user friendly and equivalently secure to use.......

In addition i sincerely appreciate a point mentioned by Gmansi of creation of jigsaw puzzle for the images but in my view point there will be listing of some particular category of images and those ranked higher in indexing system to be used as jigsaw puzzle.

as an additonal help we can use Extension Assira to make our extension smarter.

please give your valuable suggestions as we can work to improve this amazing project. :) AalekhN (talk) 22:01, 27 February 2014 (UTC)

Ahem no we can't use Asirra on Wikimedia projects because it's proprietary so it can't be self-hosted, and we can't load resources from "foreign" websites for privacy reasons. Why is why we'd need to invent a reCAPTCHA-like solution on our own if we wanted one. Nemo 20:07, 28 February 2014 (UTC)

Captcha Idea Proposal for GSOC 2014

Hello,

From last response on proposal for Multilingual, usable and effective captchas; i figured out following few solutions to the points raised by the mailing list members:

1)Captcha on the basis of selection of particular object:In this type of captcha the questions will be shown as shown here: https://commons.wikimedia.org/wiki/File:Proposal6correction.png ;other possible questions can be:"Select the images in which man is wearing sunglasses",now problem i encountered while making such captcha request is that a bot could easily use Google images or third eye to find the look for images of man wearing sunglasses from the Wikimedia Commons .....so to encounter this problem we can use a random numbering over the images and then ask user to to select the images in order of numbering for example the only answer to the sample question i provided above is :https://commons.wikimedia.org/wiki/File:Proposal7correction.png ;one of the wrong answer to the given question is:https://commons.wikimedia.org/wiki/File:Proposal8correction.png ; because it is not arranged in correct order which is 1,4,12 (in increasing numeric order).....ps:i did forgot to specify "increasing" numeric order in the question i provided but the plan is to provide the question as "Select human among the given photograph's and select it in increasing numeric order?".

Providing increasing numeric order might make bot guess wrong answer 90% of time,which is quite comparable to the current captcha system that uses OCR as provided in the given algorithm: http://www.gizmag.com/captcha-beating-ai/29559/ .

2)Ask User to click on the same image as provided:The question for this type of captcha looks like this:https://commons.wikimedia.org/wiki/File:Proposal5correction.png we can ask user to click on the options showing equivalent image to that provided in the question.For the question provided the answer is :https://commons.wikimedia.org/wiki/File:Proposal9correction.png .Ps:we ca use more than four options probably eight options since it would be harder of bot to make a guess.

3)For blind and visually impaired users:We can use and audio captcha system which ask user to select the number as it is asked in the audio .For example the visual equivalent of the audio asked by the user will be : https://commons.wikimedia.org/wiki/File:Proposal11correction.png .Now when audio asked user to select number "0" our user will use arrow key to move across different blocks like a slide show....with different voice speaking out the options and pressing enter will select the word spoken....hence verifying that the user is human.....although the above shown image is visual equivalent the actual image visible will be:https://commons.wikimedia.org/wiki/File%3AProposal4correction.png ;....to make the captcha reload we can user to by pressing key "r" which will be instructed in the audio while the captcha starts playing.

Please give your response to idea provided and also the idea is listed here:https://www.mediawiki.org/wiki/Talk:CAPTCHA

Thank You Aalekh Nigam aalekhN https://www.mediawiki.org/wiki/User:AalekhN AalekhN (talk) 10:12, 2 March 2014 (UTC)

for point no (3) got an reply on i18n Mailing List:

3) Please note that most of the blind users can't press R while

playing audio. Screen readers provide a so-called virtual cursor on

web pages, and the keys are reserved for quick navigation (H for

heading, L for list, R for region or radio button).

based upon the reply i suggested following solution to encounter it:

we should repeat the audio for two times....pause for 5 seconds then give a new captcha to the user this can help us with bot attacks we can reserve a key lets say "a" for the users to be able change the captcha if the sound is not audibleto the users.....also i came across an interesting idea in this mentoined here :http://webinsight.cs.washington.edu/projects/audiocaptchas/ .... i strongly believe the solution i suggested here can produce ease for our user AalekhN (talk) 17:46, 2 March 2014 (UTC)

MotionCAPTCHA

Take a look at the interesting idea MotionCAPTCHA.

I'm not sure that it's ready to use (at least it should work on the server side), but I think it is possible to use this technology to combat bots.

I think to successfully implement is necessary:

• Generator pictures (order to prevent reuse of a response)
• Analyzer entropy (it may be possible to find a filter's formula to separate a human entropy from a computer)
• Maybe something else, it is necessary expert opinion and brainstorming Pastakhov (talk) 04:33, 4 March 2014 (UTC)

Is there any research backing the concept? Nemo 20:22, 6 March 2014 (UTC)

I have no skills in this area.

It's just an idea to brainstorm.

Maybe it can inspire someone. Pastakhov (talk) 02:43, 7 March 2014 (UTC)

I didn't ask you to write a paper :) if you can do a Google Scholar search it would helpful. Nemo 07:04, 7 March 2014 (UTC)

Interested in this challenging project for Gsoc14

Hello, I would like to work on the CAPTCHA Project. I've already good experience in breaking weak captchas and finding bugs in them. So basically, I would suggest to introduce a new captcha design similar to re-captcha that is totally based on good image processing which could be easily implemented by some research on it. The current captcha is weak and need really some improvement and redesign. 1) For Blind people, the audio captcha should also be taken care of. 2) The design should be made in such a way the rendering doesn't eat much resources.

I would like to discuss more about it, looking forward to hear from the mentors.

Thank you. Dheerajjoshi1991 (talk) 07:31, 4 March 2014 (UTC)

Where do you plan to take the images from? Do you know Wikisource, have you tried using Proofread Page a bit and did you see how DjVu files work? Nemo 20:20, 6 March 2014 (UTC)

Proposal link for GSOC 2014

Hello, Here is the link to the proposal i prepared for GSOC 2014 : https://www.mediawiki.org/wiki/User:AalekhN/GSoC_proposal_2014

Please give your opinion and advice regarding the content of the proposal.

Thank You AalekhN (talk) 21:59, 6 March 2014 (UTC)

PHP CAPTCHA

Anything useful in there? https://github.com/dapphp/securimage They seem to have audio support. Nemo 08:39, 20 March 2014 (UTC)

This system of captcha is somewhat equivalent to the present type of captcha in use, this wont provide any additional protection from the bots also the audio captcha in use can be made more convenient by improving Screen Reader as mentioned here: http://webinsight.cs.washington.edu/projects/audiocaptchas/ AalekhN (talk) 07:16, 21 March 2014 (UTC)

visualCaptcha

https://github.com/emotionLoop/visualCaptcha : anything of use? Requires Flash it seems; shows 5 icons (like clock, woman, camera, sunglasses, key) and calls one by name asking for it to be clicked. The name could be translated but I have no idea how such a thing can work, probably most protection is given by the inability of spambots to do anything in Flash? Nemo 09:02, 20 March 2014 (UTC)

Yes this could be a good alternative i just experimented with it on my local computer and gave a thought included it on my proposal. :) AalekhN (talk) 08:49, 21 March 2014 (UTC)

Asirra broken?

«In this article we present our Cats-Dogs-Classifier, which can tell whether a given image shows a dog or a cat with an accurancy of 80%. We achieve this by reproducing the results of the paper Machine Learning Attacks Against the Asirra CAPTCHA (Philippe Golle).» http://nbviewer.ipython.org/github/Safadurimo/cats-and-dogs/blob/master/catsdogs.ipynb Nemo 09:19, 20 March 2014 (UTC)

Concept: Digitizing for Wikisource

One thing we might consider is a variation on reCaptcha's early goal: let's use WMF captchas to help digitize scanned texts for Wikisource.

Model:

• Scanned texts (dejavu?) are processed through OCR.
• OCR issues are identified (e.g. scanned text 'word' caught by spell check as misspelling, image region clipped for use in captcha)
• One of two images presented in captcha is drawn from a pool of OCR issues, the 'solution' for this image should match a spelling dictionary or fuzzy match the OCR text. Solutions are stored until a statistically significant percentage of results are exactly the same.
• The other of two images presented in captcha must match solution exactly. Amgine (talk) 04:40, 30 March 2014 (UTC)

Hand gesture CAPTCHA

Hi! Test my new CAPTCHA, please: http://gestcha.com - simple as a toy! 146.185.31.213 17:34, 10 June 2014 (UTC)

Bauernegro - What system are we using!?

I work quite extensively on several different Wikipedias, and CAPTCHAs are normally only a minor nuisance for me. I was importing templates to the Zulu Wikipedia when I was met by the CAPTCHA code BAUERNEGRO. What is this? Are we not applying any filtering at all on the words we use? This was after already being prompted to write gipsydick. This is extremely offensive, and not acceptable from a major website like Wikipedia or from MediaWiki. CFCF (talk) 13:05, 29 August 2014 (UTC)

There is a blacklist but very rough. Nemo 18:55, 29 August 2014 (UTC)

Any way to expand it to include words such as negro, gipsy and dick? CFCF (talk) 19:48, 2 September 2014 (UTC)

Sure, you can send patches: for instance I have one which does that among other things, gerrit:121255 (based on Wiktionary).

However it may take months or years for the new blacklist to take effect, as it's only used when the captcha images are regenerated. We also don't have control on what sort of dictionary WMF uses. Nemo 22:11, 2 September 2014 (UTC)

change captcha questions

This post by Nemo bis was moved on 2014-09-19. You can find it at Project:Support desk/Flow/2014/09#h-change_captcha_questions-2014-09-19T06:09:00.000Z. Nemo 08:21, 19 September 2014 (UTC)

check for dead link in section CAPTCHA#See_also

no hope, site obsolete -> http://thread.gmane.org -- Christian 🇫🇷 FR (talk) 12:09, 15 March 2023 (UTC)

Incompatibility with assistive readers

Apologies for posting here, but I can no longer post on en.wikipedia.org.

That site now has a captcha to log in. It is not compatible with any of my readers, so I can no longer log in to my account.

My IP address is dynamic, and is almost always blocked as part of a major rangeblock (which covers my mobile telco's gateway), so I cannot edit anonymously either.

If any captcha advocate reads this, please ask the English Wikipedia community why they have blocked editors less able than themselves, and what they propose to do about it. Thank you. 85.255.237.222 08:30, 28 March 2025 (UTC)