Sub-band VAD Based on Continuous Noise Estimation in Wavelet Domain

• Autorzy: Rasoolzadeh S. , Rahmani M.

Warianty tytułu

PL

Określanie szumów przy wykorzystaniu transformaty falkowej i detektora VAD

• Języki publikacji: EN

Abstrakty

EN

Voice Activity Detectors (VADs) are widely used in speech processing applications such as speech enhancement and coding. In speech enhancement systems, we use VAD in order to obtain more accurate estimation of noise. Voice activity detection methods usually work in time or frequency domains. In this paper, we propose two approaches for voice activity detection in wavelet domain for continuous noise estimation: Subband VAD and Full-band VAD. We use the correct noise/speech classification rate to show the acceptable performance of our proposed VADs. In addition, we apply sub-band VAD to a speech enhancement system. Objective and subjective measures show that Sub-band VAD outperforms Fullband VAD in noise reduction applications.

PL

Analizowano detektor aktywności głosowej VAD w zastosowaniu do poprawy jakości dźwięku mowy. Celem było określenie szumów towarzyszących. Zastosowano transformatę falkową do analizy szumów.

Słowa kluczowe

EN

website; quality metrics; evaluation

PL

dźwięk mowy; detektor głosu; transformata falkowa

• Wydawca: Wydawnictwo SIGMA-NOT
• Czasopismo: Przegląd Elektrotechniczny
• Rocznik: 2012
• Tom: R. 88, nr 2
• Strony: 311--314
• Opis fizyczny: Bibliogr. 11 poz., rys., tab., wykr.

Twórcy

autor

Rasoolzadeh S.

autor

Rahmani M.

• Department of Computer Engineering, Malayer Branch, Islamic Azad University, Malayer, Iran,

Bibliografia

• [1] M. Rahmani, M. Mohammadi, A. Akbari, "Back ground noise control for speech enhancement," in Proc.14th ICEE, IRAN, Tehran, 2006.
• [2] F. Beritelli, S. Casale, A. Cavallaero, "A robust voice activity detector for wireless communication using soft computing," IEEE Journal, Vol. 16, Issue. 9, pp 1818-1829, December 1998.
• [3] P. Renevey, A. Drygajlo, "Entropy Based Voice Activity Detection in Very Noisy Conditions," in Proc. Eurospeech 2001, Aalborgh 2001.
• [4] V. Gilg, C.Beaugeant, M.Schonle, "Methodolog for the design of a robust voice activity detector for speech enhancement," International Workshop on Acoustic echo and Noise Control (IWAEN 2003)
• [5] Y. Kida, T. Kawahara," Voice Activity Detection based on Optimally Weighted Combination of Multiple Features," in Proc. Eurospeech , pp. 2621-2624,2005.
• [6] A. Grossman, R. Martinet., J. Morlet , "Reading and understanding continuous wavelet transform," Springer-Verlag, pp. 2-20, 1989
• [7] I. Yann, S. Ngee, C. Kiat, "Wavelet for speech denoising, " IEEE TENCON vol.2, pp. 479 – 482. 1997.
• [8] P. Sovka, P. Pollak, J. Kybic, "Extended Spectral Subtraction," in Proc. EUSIPCO 1996, Trieste, Italy, September. 1996.
• [9] F. Beritelli, S. Casale, G. Ruggeri, S. Serrano, "Performance evaluation and comparison of G.729/AMR/fuzzy voice activity detectors," IEEE Signal Processing Letters,Vol. 9 , Issue 3 , pp.85 – 88, March 2002.
• [10] C. JIA, B. XU, "An Improved Entropy-Based Endpoint Detection Algorithm," in Proc. ICSLP 2002, Beijing 2002
• [11] E. Kosmides, E. Dermatas, G. Kokkinakis, "Stochastic endpoint detection in noisy speech," SPECOM Workshop, pp. 109-114, 1997.