# Extracting Audio

# AWS WAF CAPTCHA

Currently, the only CAPTCHA for which we use speech recognition is AWS WAF (Amazon Web Services Web Application Firewall) CAPTCHA. Extracting the base64 audio data that's needed for the recognition is trivial.

AWS sends us an audio element already loaded with base64 audio data, so extracting this is as simple as:

const $audio = document.querySelector('audio');
const audio_data = $audio.src.replace('data:audio/aac;base64,', '');

Thanks Jeff Bezos!

AWS WAF CAPTCHA Recognition API

../../recognition/awscaptcha/