ePrints@IIScePrints@IISc Home | About | Browse | Latest Additions | Advanced Search | Contact | Help

Gammatone Wavelet Cepstral Coefficients for Robust Speech Recognition

Adiga, Aniruddha and Magimai-Doss, Mathew and Seelamantula, Chandra Sekhar (2013) Gammatone Wavelet Cepstral Coefficients for Robust Speech Recognition. In: IEEE International Conference of Region 10 (TENCON), OCT 22-25, 2013, Xian, PEOPLES R CHINA.

[img] PDF
ieee_int_con_rec_2013.pdf - Published Version
Restricted to Registered users only

Download (639Kb) | Request a copy
Official URL: http://dx.doi.org/10.1109/TENCON.2013.6718948

Abstract

We develop noise robust features using Gammatone wavelets derived from the popular Gammatone functions. These wavelets incorporate the characteristics of human peripheral auditory systems, in particular the spatially-varying frequency response of the basilar membrane. We refer to the new features as Gammatone Wavelet Cepstral Coefficients (GWCC). The procedure involved in extracting GWCC from a speech signal is similar to that of the conventional Mel-Frequency Cepstral Coefficients (MFCC) technique, with the difference being in the type of filterbank used. We replace the conventional mel filterbank in MFCC with a Gammatone wavelet filterbank, which we construct using Gammatone wavelets. We also explore the effect of Gammatone filterbank based features (Gammatone Cepstral Coefficients (GCC)) for robust speech recognition. On AURORA 2 database, a comparison of GWCCs and GCCs with MFCCs shows that Gammatone based features yield a better recognition performance at low SNRs.

Item Type: Conference Proceedings
Related URLs:
Additional Information: copyright for this article belongs to IEEE, 345 E 47TH ST, NEW YORK, NY 10017 USA
Keywords: Gammatone wavelets; Auditory modeling; Cepstral coefficients; Speech recognition
Department/Centre: Division of Electrical Sciences > Electrical Engineering
Date Deposited: 09 Jun 2014 09:54
Last Modified: 09 Jun 2014 09:54
URI: http://eprints.iisc.ernet.in/id/eprint/49235

Actions (login required)

View Item View Item