Skip to content

Speech Enhancement Metrics (PESQ, CSIG, CBAK, COVL)

Notifications You must be signed in to change notification settings

nglehuy/semetrics

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Speech Enhancement Metrics

This implementation of Matlab was originally taken from https://ecs.utdallas.edu/loizou/speech/software.htm

REQUIREMENTS

Octave must be installed

apt install octave
pip install oct2py Cython # install Cython before installing pesq
pip install pesq

COMPOSITE MEASURE

Usage:

[Csig,Cbak,Covl,segSNR]=composite(cleanfile.wav,enhanced.wav)

where Csig is the predicted rating of speech distortion, Cbak is the predicted rating of background distortion, Covl is the predicted rating of overall quality.

You may run example files included in the zip file. In MATLAB, type:

>> [c,b,o,s]=composite('sp09.wav','enhanced_logmmse.wav')

c = 3.3050

b = 2.6160

o = 2.7133

s = 3.9917

where sp09.wav is the clean file and enhanced_logmmse.wav is the enhanced file.

References

MATLAB Source: https://ecs.utdallas.edu/loizou/speech/software.htm

@ARTICLE{4389058,
  author={Y. {Hu} and P. C. {Loizou}},
  journal={IEEE Transactions on Audio, Speech, and Language Processing}, 
  title={Evaluation of Objective Quality Measures for Speech Enhancement}, 
  year={2008},
  volume={16},
  number={1},
  pages={229-238},
}

Releases

No releases published

Packages

No packages published