Skip to content

Latest commit

 

History

History

rVAD2.0

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
LICENSE
LICENSE
 
 
README.TXT
README.TXT
 
 
aurora2read.m
aurora2read.m
 
 
enframe.m
enframe.m
 
 
estnoiseg.m
estnoiseg.m
 
 
estnoisem.m
estnoisem.m
 
 
estnoisem_noiseseg.m
estnoisem_noiseseg.m
 
 
filtbankm.m
filtbankm.m
 
 
findpeaks.m
findpeaks.m
 
 
fxpefac.m
fxpefac.m
 
 
gaussmixp.m
gaussmixp.m
 
 
irfft.m
irfft.m
 
 
pitchblockdetect.m
pitchblockdetect.m
 
 
pitchestm.m
pitchestm.m
 
 
rfft.m
rfft.m
 
 
sflux.m
sflux.m
 
 
snre_highenergy.m
snre_highenergy.m
 
 
snre_vad.m
snre_vad.m
 
 
specsub.m
specsub.m
 
 
specsub_noiseseg_lfn.m
specsub_noiseseg_lfn.m
 
 
spgrambw.m
spgrambw.m
 
 
stdspectrum.m
stdspectrum.m
 
 
vad.m
vad.m
 
 
vadbatch_1folder_diffpathes.m
vadbatch_1folder_diffpathes.m
 
 
voicebox.m
voicebox.m
 
 
winenvar.m
winenvar.m
 
 

README.TXT

Noise-robust voice activity detection algorithm (rVAD).
Version 2.0

28 Nov 2017, Zheng-Hua Tan

Usage: vad(finwav, fvad) 
       vad(finwav, fvad, opts) 
       vad(finwav, fvad, opts, vadThres). 

where finwav is the input WAVE file path and name, fvad is the output VAD file path and name, opts can be 0 for using pitch (default option) or 1 for using flatness (significantly faster at the cost of slightly reduced accuracy), and finally vadThres is the threshold for VAD. Refer to vad.m for more detailed explanation. 

The code has been tested on Matlab R2016a.

Refs:
[1] Z.-H. Tan, A.k. Sarkara and N. Dehak, "rVAD: an unsupervised segment-based robust voice activity detection method," Computer Speech and Language, vol. 59, pp. 1-21, 2020. 
[2] Z.-H. Tan and B. Lindberg, "Low-complexity variable frame rate analysis for speech recognition and voice activity detection,” IEEE Journal of Selected Topics in Signal Processing, vol. 4, no. 5, pp. 798-807, 2010.

Contact:
  Prof Zheng-Hua Tan
  Aalborg University, Denmark
  [email protected]
  https://vbn.aau.dk/en/persons/107665