Text-independent speaker recognition

• Autorzy: Bobulski J.
• Języki publikacji: EN

Abstrakty

EN

In this paper, text-independent speaker recognition method based on Wavelet Transform and melcepstrum is presented. The results of experiments point the best parameters of Wavelet Transform for speaker identification, and can be useful for design speaker identification systems. This kind method of person identification is useful in services such as banking by telephone, access authorization to resources and for forensic purpose.

Słowa kluczowe

EN

speaker identification; voice recognition

PL

identyfikacja mówcy; rozpoznawanie mowy

• Wydawca: Wydawnictwo Politechniki Częstochowskiej
• Czasopismo: Computing, Multimedia and Intelligent Techniques
• Rocznik: 2007
• Tom: Vol. 3, nr 1
• Strony: 101--107
• Opis fizyczny: Bibliogr. 6 poz., rys., tab.

Twórcy

autor

Bobulski J.

• Czestochowa University of Technology, Institute of Computer and Information Sciences, ul. Dąbrowskiego 73, 42-200 Częstochowa, Poland,

Bibliografia

• [1] Furui S., Survey of the State of the Art in Human Language Technology, Tokyo, Japan, http://cslu.cse.ogi.edu/HLTsurvey/HLTsurvey.html, 1996.
• [2] Aboufadel E., A Wavelets Approach to Voice Recognition, http://www.gvsu.edu/math/wavelets/student_work/Hoekstra/voice_recognition.htm -29 November, 2001.
• [3] Misiti M, Misiti Y., OppenheimG., Poggi l.-M.,Wavelet Toolbox, MathWorks, 1998.
• [4] Augustyniak P., Transformacje falkowe w zastosowaniach elektrodiagnostycznych, Uczelniane Wydawnictwa Naukowo-Dydaktyczne, Kraków 2003 (in Polish).
• [5] Kinnunen T., Features for Automatic Text-Independent Speaker Recognition, http://www.cs.joensuu.fi/pages/pums/public_results/2004_PhLic_Kinnunen_Tomi.pdf -21st December 2003.
• [6] Kubanek M., The Method of Audio-Visual Polish Speech Recognition Based on Hidden Markov Models, PhD Thesis, Częstochowa University of Technology, Częstochowa, Poland, 2005 (in Polish).