Hi am Ahmad i would like to get details on mfcc 39 matlab code feature extraction ..My friend Justin said mfcc 39 matlab code feature extraction will be available here and now i am living at ......... and i last studied in the college/school ......... and now am doing ....i need help
My email is : ahmad.darkhali[at]gmail.com
Posts: 14,118
Threads: 61
Joined: Oct 2014
Given that Mel-frequency Cepstral Coefficients, the other really popular voice feature, involves almost the same processing steps, I decided to make an implementation for them too, using the same blocks as much as possible. Voice recognition is a typical example. Through more than 30 years of recognition research, many representations of different characteristics of the speech signal have been suggested and tested. The most popular function representation currently used is Mel-frequency Cepstral Coefficients or MFCC.
Another popular representation of the speech feature is known as RASTA-PLP, an acronym for Relative Spectral Transformation - Perceptual Linear Prediction. The PLP was originally proposed by Hynek Hermansky as a way to deform the spectra to minimize differences between speakers while preserving important speech information [Herm90]. RASTA is a separate technique that applies a band-pass filter to the energy in each frequency sub-band to smoothen short-term noise variations and eliminate any constant displacement resulting from static spectral coloration in the voice channel, p. from a telephone line.