FeatureSet: plp methods

Syntax:

   <feature set>  auditory <feature> <source> [-nf <filterN>] 
   <feature set>  postaud  <feature> <source>
   <feature set>  plp      <feature> <source> [-o <order>] [-n <filterN>] 

Perceptual Linear Prediction (PLP) is broken into three steps to give the possibility of additional processing (like RASTA). For more information see [1] H.Hermansky, "Perceptual linear predictive analysis for speech", J. Acoust. Soc. Amer., pp. 1738-1752, 1990.

Optional flags:

-nf <filterN>

-n <filterN>

Number of filters in the filterbank.

-o <order>

Order of ceptral analysis. The output dimension will be <order>+1.