You are here

Acoustic features

A recognition system typically uses some auditory-based representations of the speech signal: well-known features are Mel Frequency Cepstral Coefficients (MFCC) or Perceptual Linear Prediction (PLP).