pyannote-shennong-features

A feature extractor for Pyannote that is based on the Shennong library

Installation

This library depends on Shennong, which is available on conda. Thus, make sure you're in a conda environment where pyannote-audio has been installed already and run

conda install -c coml shennong

Then you can pip-install the library in a regular way:

pip install pyannote.features.shennong

Pyannote Configuration

To use features from this library, this is the kind of configuration you should add your config.yml file:

feature_extraction:
   name: pyannote.features.ShennongMfcc
   params:
       coefs: 19
       e: True
       D: True
       DD: True
       mfccWindowType: 'hanning'
       mfccLowFreq: 20
       mfccHighFreq: -100 # Real value will be (f_nyquist - 100)
       with_pitch: True
       with_cmvn: False
       duration: 0.025
       step: 0.010
       sample_rate: 16000
       pitchFmin: 20
       pitchFmax: 500