Amphion Vevo: Open Source Instant Voice Cloning (Accents Too!)

[@guitarmeetsscience]

You deserve a bigger following - Love these videos!

[@ParvathyKapoor]

pinokio install?

[@ChameleonAI]

Great idea. I just published github.com/chameleon-ai/vevo-pinokio

[@ParvathyKapoor]

@ChameleonAI thanks boss 🙏

[@kowy_m]

it's confusing detecting u'r voice(does it exits !?), place a beep or other to know that's narrative

[@ChameleonAI]

You mean my "real" voice? I always use a voice changer, so in a sense my real voice doesn't exist.

[@Maxume]

Could you maybe explain what benefit this offers over F5-TTS? The quality doesn't seem to be better but the VRAM requirement is double or more than that of F5.

[@ChameleonAI]

There is a TTS option but this is primarily voice-to-voice, so it's a different application. Unless I missed something, F5 doesn't have an sts mode.

[@Maxume]

@@ChameleonAI Oh, sorry. I thought it was TTS. It makes sense now. Thanks

[@JuicyBenji]

Nice, is this real time as well or only recordings?

[@ChameleonAI]

Only recordings. I didn't see a way of feeding in a stream real-time with the code they have right now, though it may be theoretically possible because it takes less time to run the conversion than the duration of the clip (on my machine anyway).

[@numberonedad]

that american guy is missing an alveolar flap in "water" americans say wadder not waTer

[@ChameleonAI]

Probably over-annunciating because he was doing a reading. I do that myself sometimes.

[@mgga24]

sent a discord dm 😶‍🌫