Paper Info
Title
A Novel Windowing Technique for Efficient Computation of MFCC for Speaker Recognition
Abstract
In this letter, we propose a novel family of windowing technique to compute mel frequency cepstral coefficient (MFCC) for automatic speaker recognition from speech. The proposed method is based on fundamental property of discrete time Fourier transform (DTFT) related to differentiation in frequency domain. Classical windowing scheme such as Hamming window is modified to obtain derivatives of discrete time Fourier transform coefficients. It is mathematically shown that this technique takes into account slope of power spectrum and phase information. Speaker recognition systems based on our proposed family of window functions are shown to attain substantial and consistent performance improvement over baseline single tapered Hamming window as well as recently proposed multitaper windowing technique.
Year
DOI
Venue
2012
10.1109/LSP.2012.2235067
Signal Processing Letters, IEEE
Keywords
Field
DocType
cepstral analysis,differentiation,discrete Fourier transforms,frequency-domain analysis,speaker recognition,DTFT,MFCC,automatic speaker recognition,baseline single tapered Hamming window,classical windowing scheme,differentiation,discrete time Fourier transform coefficients,frequency domain,fundamental property,mel frequency cepstral coefficient,multitaper windowing technique,performance improvement,phase information,power spectrum,speaker recognition systems,window functions,Differentiation in frequency,mel-frequency cepstral coefficients (MFCC),power spectrum estimation,speaker recognition,tapered window
Discrete-time Fourier transform,Frequency domain,Mel-frequency cepstrum,Pattern recognition,Multitaper,Computer science,Speech recognition,Spectral density,Speaker recognition,Artificial intelligence,Window function,Computation
Journal
Volume
Issue
ISSN
20
2
1070-9908
Citations 
PageRank 
References 
13
0.69
8
Authors
2
Name
Order
Citations
PageRank
M. Sahidullah1130.69
Goutam Saha225523.17