CAPTCHAs are tests that are used to detect spammers and automated processes masquerading as humans. To access a resource, the user must solve a problem that cannot easily be solved by a computer. The most common form is the Visual CAPTCHA, which consists of some distorted text:
The human who wishes to access the protected resource must correctly identify the text that is displayed. Since users with poor vision cannot solve these problems, and alternative called the Audio CAPTCHA has been developed. It is often combined in the same interface as a Visual CAPTCHA, as in the following example, reCAPTCHA:
In this kind of test, the user must listen to a garbled voice and identify the numbers or words that are being spoken.
Audio CAPTCHAs are more difficult than visual CAPTCHAs, especially for blind users who use screen-reader programs. These programs tend to talk over the audio CAPTCHA as the user navigates between the audio playback controls and answer box, frustrating the user's ability to hear the CAPTCHA. Moreover, the blind user must memorize a list of random numbers or words, stretching the limits of human short-term memory. Even sighted users have less than 50% success rate for the most common Audio CAPTCHAs, due to the intrinsic difficulty of interpreting the noisy sound file.
Rather than creating a new type of CAPTCHA, the approach here is to improve the interface of preexisting Audio CAPTCHAs, by allowing the user to control the playback of the sound without leaving the answer box. The following controls are available to the user:
With this approach, there is no need for separate playback controls.