Technical Program

Paper Detail

Paper:SP-L4.6
Session:Higher-Level Knowledge in Speaker Recognition
Time:Wednesday, May 19, 17:10 - 17:30
Presentation: Lecture
Topic: Speech Processing: Speaker Recognition
Title: IMPROVEMENT OF SPEAKER RECOGNITION BY COMBINING RESIDUAL AND PROSODIC FEATURES WITH ACOUSTIC FEATURES
Authors: Shi-Han Chen; National Tsing Hua University 
 Hsiao-Chuan Wang; National Tsing Hua University 
Abstract: When a speech signal is encoded in some low bit-rate coding formats, it becomes more difficult in distinguishing the speaker identities. This paper investigates the codec effect to acoustic and prosodic features. A new representation of prosodic features is introduced. The inclusion of residual features is based on LDA algorithm. By combining residual and prosodic features with the acoustic features, we can improve the performance of speaker recognition system. A series of experiments is performed with coded speech affected by G.729A and GSM codec process to demonstrate the effectiveness of our proposed method.
 
           Back


Home -||- Organizing Committee -||- Technical Committee -||- Technical Program -||- Plenaries
Paper Submission -||- Special Sessions -||- ITT -||- Paper Review -||- Exhibits -||- Tutorials
Information -||- Registration -||- Travel Insurance -||- Housing -||- Workshops

©2015 Conference Management Services, Inc. -||- email: webmaster@icassp2004.org -||- Last updated Wednesday, April 07, 2004