Language Independent Speaker Identification

Automatic speaker identification (ASI) and Automatic speaker verification (ASV) are probably the most natural and economical methods for solving the problems of unauthorized use of computer and communications systems and multilevel access control. This paper presents a method for speaker identification, independent of language spoken. This paper uses combination of both, pitch frequency and speaker specific vocal tract information to separate one speaker from other. The identification process goes through future extraction, spectral analysis, Welch analysis and pitch detection or estimating fundamental frequency using harmonic product spectrum (HPS) algorithm [3].. The speaker specific vocal tract information is mainly represented by mel-frequency cepstrum coefficients(MFCCs) [5]. The proposed method uses MFCCs as additional features along with pitch histograms (histogram of pitch frequency values), since these two are complementary features, they give better identification [4].