Create a sample audio captcha

[hrj]

We could use a Text-to-speech (TTS) system to generate an audio sample, and store it using a well known format, like Opus, etc.

The stretch goal would be to do it in such a way that voice recognition software can't recognise it. So we might have to add some distortions to the audio.

[myleskeeffe]

I don’t know if you necessarily need to add Voice-to-Text preventions - as a VTT translation would use a fair amount of machine power - so anything more then just a few bots would be way to slow and taxing on a spammer id hope hehe

[hrj]

@myleskeeffe Good point. Maybe just adding a bit of white noise to the clip might be more than enough, and push the cost for the bots even higher, with very little cost on our end.

[myleskeeffe]

Sounds like a good idea - and depending which TTS library is used - could use multiple ‘accents’ to prevent bots just recognising how a word is said (I mean if we had a dictionary of like 30k words - it’d be pretty unlikely for someone to make a bot that matches their own library of the TTS recordings (as that’d grow to a fair few gbs) - and the bit of white noise would prevent them from just running a direct match against such a library anyways)

[oopsliu]

Hi, is there an estimated timeframe for supporting the audio captcha? We are interested in integrating LibreCaptcha but accessibility is a hard requirement for our website.

[hrj]

Hi

We are currently not actively working on the project due to constraints.

Are people interested in sponsoring work on this and other issues? If yes, I might be able to find folks or spend time on it myself.

React to this comment with a 👍🏽 if you or your organisation are willing to sponsor.

[oopsliu]

Thanks @hrj , I need to talk about this within my organization regarding sponsorship.