Convert audio signal to MFCC

Hiya,

Is it possible to convert an audio signal (either live microphone input or pre-recorded files) to mel frequency coefficients (MFCC) in vvvv? I would like to extract 13 MFCCs and send them via OSC to Wekinator for some simple AI speech recognition.

There’s an openFrameworks example for this on the Wekinator website (Example code | Wekinator) but I’d much prefer to have a vvvv patch that handles the conversion. Anyone here tried this before?

If you have basic coding skills it should be relatively easy to get it using this library:

You can implement that for vaudio or try the VL.NewAudio contribution.

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.