Using Larynx for text to speech?

[aladdin]

I’m not sure if you’ve seen this already, but the Rhasspy project has been creating and using a fork of Mozilla Deepspeech called Larynx. They’ve improved the resource usage, so it’s able to run effectively on a RaspberryPi 4 while still sounding good. It sounds like the next version will be even more resource efficient.

It would be amazing to see Genie be able to use this, effectively wiping out half of the Google dependencies and improving privacy even further.

[aladdin]

There’s quite a nice video demo / breakdown at:

Larynx Demo

[sileixu]

Thank you a lot for the information. One thing to correct here, we are using Microsoft’s service, not Google’s, but yeah, it’s definitely great to see an open-source option available. We will look into switching to that.