Group Delay를 이용한 GMM기반의 성별 인식 알고리즘 = GMM-Based Gender Identification Employing Group Delay

국문 초록 (Abstract)

본 논문은 Group Delay(GD)를 이용한 음성신호 기반의 효과적인 성별인식 시스템을 제안한다. 일반적인 음성 인식과 관련된 시스템에서 사용되는 특징들은 위상에 관한 정보를 제거한 크기만의 정보를 이용하여 구성한다. 본 연구에서는 위상에 관한 정보를 토대로 유도되어 지는 GD의 성별에 따른 특징을 알아보고, 보다 향상된 성별인식을 위해 MFCC(Mel-frequency cepstral coefficient), LPC(linear predictive coding) 계수, 반사계수(reflection coefficient) 그리고 포만트(formant)등과 같은 크기 정보와 GD를 이용한 결합 특징 벡터를 적용하였다. 실험을 통해 성별에 따른 GD의 특징을 확인할 수 있었고, 이를 이용한 제안된 특징 벡터를 사용했을 때 우수한 인식 성능을 얻을 수 있었다.

다국어 초록 (Multilingual Abstract)

We propose an effective voice-based gender identification using group delay(GD) Generally, features for speech recognition are composed of magnitude information rather than phase information. In our approach, we address a difference between male and female for GD which is a derivative of the Fourier transform phase. Also, we propose a novel way to incorporate the features fusion scheme based on a combination of GD and magnitude information such as mel-frequency cepstral coefficients(MFCC), linear predictive coding (LPC) coefficients, reflection coefficients and formant. The experimental results indicate that GD is effective in discriminating gender and the performance is significantly improved when the proposed feature fusion technique is applied.

참고문헌 (Reference)

1 이계환, "음성신호 기반의 성별인식을 위한 Support vector machines의 적용" 26 (26): 75-79, 2007

2 A, "” International Conference on Spoken Language Processing" -898, 1992

3 A, "” International Conference on Spoken Language Processing" -898, 1992

4 A, "“The importance of phase in signals" -541,

5 A, "“Significance of group delay functions in spectrum estimation ” IEEE Transactions on Signal Processing" 2281-2289, 1992

6 Krishnan, "“Significance of group delay function in signal reconstruction from spectral magnitude or phase” IEEE Transactions on Acoustics" 610-623, 1984

7 Schroeder, "“Models of hearing" 1332-1350, 1975

8 A, "“Formant extraction from phase using weighted group delay function" (v) : 1609-1611, 1989

9 A, "“Formant extraction from group delay function" 209-221, 1991

10 Davis, "“Comparison of parametric representations for monosyllabic word recognition in continuously spoken sentences ” IEEE Transactions on Acoustics" 357-366, 1980

11 “Perceptual linear predictive, "analysis of speech ” Journal of Acoustic Society of America" 1738-1752, 1990

12 H. Harb, "Voice-based gender identification in multimedia applications" 24 : 179-198, 2005

13 K. K. Paliwal, "Usefulness of phase spectrum in human speech perception" 2117-2120, 2003

14 N. S. Kim, "Spectral enhancement based on global soft decision" 7 (7): 108-110, 2000

15 S. Ramamohan, "Sinusoidal model-based analysis and classification of stressed speech" 14 (14): 737-746, 2006

16 K. K. Paliwal, "On the usefulness of STFT phase spectrum in human listening tests" 17 (17): 578-616, 2007

17 B. Yegnanarayana, "Non-spectral features for speech processing" 2006

18 E. S. Parris, "Language independent gender identification" 2 : 685-688, 1996

19 L. D. Alsteris, "Importance of window shape for phase-only reconstruction of speech" 1 : I-573-576, 2004

20 H. Harb, "Gender identification using a general audio classifier" 2 : 733-736, 2003

21 L. D. Alsteris, "Further intelligibility results from human listening tests using the short-time phase spectrum" 48 (48): 727-736, 2006

22 L. Liu, "Effects of phase on the perception of intervocalic stop consonants" 22 (22): 403-417, 1997

23 G. Xuan, "EM algorithm of gaussian mixture model and hidden Markov model" 1 : 145-148, 2001

24 R. Smits, "Determination of instants of significant excitation in speech using group delay function" 3 (3): 325-333, 1995

25 S. Slomka, "Automatic gender identification optimised for language independence" 1 : 145-148, 1997