Import Piper / VITS as iOS Custom System Voices for TTS using AVSpeechSynthesisProviderAudioUnit Extension in Swift

[S-Ali-Zaidi]

In recent iOS updates, Apple now allows for any speech synthesizers brought into iOS via an App to become a part of the custom options for the iOS system voices for TTS, using the AVSpeechSynthesisProviderVoice and AVSpeechSynthesisProviderAudioUnit:

See:

• Creating a custom speech synthesizer
• Extend Speech Synthesis with personal and custom voices

This functionality has been implemented by the official espeak-ng iOS app, you can see how it was done in extensions directory of their iOS app source code

There is also a repo that has built out a similar functionality around espeak-ng and piper for iOS here

Wondering if we can see this implemented within Sherpa-onnx too?

[csukuangfj]

Would you like to contribute?

We have done that for Android.

[willwade]

I think you’d be better forking that iOS piper code. I built something similar and scrapped it because of that code. As piper is vits we should be able to largely use the same thing. We need this too