Identifying Immediate Family Relationship by Extraction of Parametric Models and Statistical Features of Cepstrum of Speech Signal
[1]
Morteza Zangeneh Soroush, Department of Bioelectric Engineering, College of Biomedical Engineering, Tehran Science and Research Branch, Islamic Azad University, Tehran, Iran.
[2]
Keivan Maghooli, Department of Bioelectric Engineering, College of Biomedical Engineering, Tehran Science and Research Branch, Islamic Azad University, Tehran, Iran.
[3]
Zahra Abbasvandi, Department of Bioelectric Engineering, College of Biomedical Engineering, Tehran Science and Research Branch, Islamic Azad University, Tehran, Iran.
[4]
Sara Bagherzadeh, Department of Bioelectric Engineering, College of Biomedical Engineering, Tehran Science and Research Branch, Islamic Azad University, Tehran, Iran.
[5]
Mohammad Mohammadi Erbati, Department of Biomedical Engineering, Qazvin Branch, Islamic Azad University, Qazvin, Iran.
[6]
Niusha Saboonchi, Department of Bioelectric Engineering, College of Biomedical Engineering, Tehran Science and Research Branch, Islamic Azad University, Tehran, Iran.
This paper investigates the possibility of identifying the immediate relationship between family members using the voice of the members. The purpose of this article is to decide whether there are common features in the voice of family members or not. In other words, is it possible to identify the relationship between the immediate family members from their voice? This article is based on identifying the immediate family relationship through processing vocal signals. After taking a vocal signal from a rich phonetic text and doing the preprocessing phase, there will be an attempt to design an immediate family identification system with the aid of statistic features of cepstrum and the parameters in Autoregressive-moving-average model and Autoregressive model. In order to evaluate proposed system we used two different methods and computed False Acceptance Rate, False Rejection Rate and accuracy. The accuracy of proposed system for first and second method was 40% and 65%, respectively. Results show proposed system was good in identifying immediate family relationship.
Family, Speaker Identification, Biometrics, Parametric Models, Cepstrum, Speech Processing
[1]
B. S. Atal, “Effectiveness of Linear Prediction Characteristics of the Speech Wave for Automatic Speaker Identification and Verification” Journal of the Acoustical Society of America. vol 55,pp.1304–1312. 1974.
[2]
Y.Bennani, “Text-Independent Talker IdentificationSystem Combining Connectionist and Conventional Models” IEEE-SP Workshop on Neural Networks forSignal Processing, IEEE Service Center Press. pp. 131-138. 1992.
[3]
B. P. Bogert, M. J. R. Healy and J.W. Tukey, “The frequencyAnalysis of Time Series for Echoes: Cepstrum, Pseudo Auto-covariance, Cross-Cepstrum and Saphe Cracking” Proceedings of the Symposium on Time Series Analysis" (M. Rosenblatt, Ed) Chapter 15, Wiley, New York, pp.209-243. 1963.
[4]
G. Box, G.M, Jenkins and G.C. Reinsel, “Time Series Analysis: Forecasting and Control”, third ed. Prentice-Hall.1994.
[5]
J.Campbell, “Speaker Recognition: A Tutorial”. Proc. IEEE. vol 85(9), 1437-1462. 1997
[6]
S. Fine, J. Navratil, and R.Gopinatth, “A HybridGMM/SVM Approach to Speaker Identification”. IEEE Int. Conf. on Acoustics, Speech, and Signal Processing. 2001.
[7]
S. Fine, J. Navratil and R. Gopinatth, “EnhancingGMM Scores Using SVM ‘Hints”, Seventh European Conference on Speech Communication and Technology. 2001
[8]
S. Furui, “Cepstral Analysis Technique for Automatic Speaker Verification”. IEEE Transaction ASSP.29, pp.254–272. 1981.
[9]
E. J. Hannan, M. Deistler, “Statistical theory of linear systems”. Wiley series in probability mathematical statistics. John Wiley and Sons, New York, pp.227.1988.
[10]
M.M. Homayoun, H. Razazan,”Journal of Computer Society of Iran”. Vol 3 (1), pp. 33-43.2005.
[11]
W.C.Hsu, J.N.Sun. “The Effectiveness of Linear Prediction Residual to the Verification of Voiceprint and the Recognition of Chinese Tone” IEEE Int Symposium on Multimedia (ISM). pp. 353-356. 2010.
[12]
J.D. Markel, S.B. Davis, “Text IndependentSpeaker Recognition from a Large Linguistically Unconstrained Time Spaced Data Base”. IEEETransaction on ASSP, vol 27(1), pp.74–82. 1979.
[13]
M.Sh. Moin and R. Boostani, “Comparing SVM, GMM and HMM methods in speaker authentication”, 11th conference on electrical engineering, Iran, pp. 286-293. 2003.
[14]
M. Norton and D. Karczub. Fundamentals of Noise and Vibration Analysis for Engineers. Cambridge University Press. 2003.
[15]
L. Wang, K. Chen, and H.S.Chi, “Towards BetterCapturing Inter-Speaker Information by Active Learning for Speaker Identification” International Joint Conference on Neural Networks (IJCNN).4, pp.2975-2980.
