An approach to speaker identification using acoustic-feature planes

  • Authors:

  • Hossein Marvi

  • Affiliations:

  • Shahrood University of Technology, Faculty of Electrical Engineering and Robotic, Shahrood, Iran

  • Venue:
  • DNCOCO'06 Proceedings of the 5th WSEAS international conference on Data networks, communications and computers
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

Speaker recognition is the use of a machine to recognise a person from a spoken phrase. Different features have been investigated in speaker recognition system. Popular features are linear predition cepstral coeffficient (LPCC) and mel frequency cepstral coefficints (MFCC) which both measure the short-term spectral envelope. This paper suggests a new feature for speaker recogntion based of time-frequency features using acoustic features planes. Experimental resuls demonstrate that the proposed method gives higher identifiction rate than LPCC and MFCC.