Optimization of Gabor Features for Text-Independent Speaker Identification

V. Mildner, S. Goetze, Karl-Dirk-Kammeyer [Unknown], A. Mertins

Abstract

For text-independent speaker identification a prominent combination is to use Gaussian mixture models (GMM) for classification while relying on Mel-frequency cepstral coefficients (MFCC) as features. To take temporal information into account the time difference of features of adjacent speech frames are appended to the initial features. In this paper we investigate the applicability of spectro-temporal features obtained from Gabor-filters and present an algorithm for optimizing the possible parameters. Simulation results on a database show that spectro-temporal features achieve higher recognition rates than purely temporal features for clean speech as well as for disturbed speech.
Original languageEnglish
Title of host publication2007 IEEE International Symposium on Circuits and Systems
Number of pages4
PublisherIEEE
Publication date01.05.2007
Pages3932-3935
ISBN (Print)1-4244-0920-9
ISBN (Electronic)1-4244-0921-7
DOIs
Publication statusPublished - 01.05.2007
Event2007 IEEE International Symposium on Circuits and Systems - New Orleans, United States
Duration: 27.05.200730.05.2007
Conference number: 70268

Fingerprint

Dive into the research topics of 'Optimization of Gabor Features for Text-Independent Speaker Identification'. Together they form a unique fingerprint.

Cite this