Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Does this package work alongside Vivox? #79

Open
Vivraan opened this issue Mar 19, 2024 · 2 comments
Open

Does this package work alongside Vivox? #79

Vivraan opened this issue Mar 19, 2024 · 2 comments
Labels
question Further information is requested

Comments

@Vivraan
Copy link

Vivraan commented Mar 19, 2024

I've set up Vivox alongside a multiplayer setup in my project, which is being deployed to Quest 2 hardware. From what I understand, Vivox uses its own libraries to access the device microphone, which can then be tested in echo and group voice channels.

My question is: does this interfere in any way with how whisper.unity handles microphone usage requests (through Unity's Microphone API?)

Additionally, I haven't gotten it to work in the Quest 2 headsets.

@Macoron
Copy link
Owner

Macoron commented Mar 24, 2024

whisper.unity uses custom MicrophoneRecord script to make circular recording and run VAD. If Vivox uses their own custom logic for working with a microphone - you would need to rewrite that to make them work together with whisper.

Alternatively, whisper can work with an AudioClip. If Vivox can provide you final AudioClip - you can transcribe it and doesn't interact with a microphone directly. Check how to transcribe audio clips in samples.

As for running on Quest 2, it should work, but very slow. Check #47

@Macoron Macoron added the question Further information is requested label Mar 24, 2024
@Vivraan
Copy link
Author

Vivraan commented Mar 24, 2024

Yes, I only just saw zhouwg/kantv#64, which shows somewhat fast inference on Android using the original whisper.cpp repo.

As for the Vivox integration, Vivox supports passing its audio outputs through Unity, so it's possible to use OnAudioFilterRead to directly pass in float arrays of the audio data to whisper.unity. I get terrible performance, though.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
question Further information is requested
Projects
None yet
Development

No branches or pull requests

2 participants