Abstract：The spectral subtraction speech enhancement is extensively used due to its simplicity and easy to implement. The principle of this method is to subtract the estimated magnitude of the noise from the magnitude of the noisy signal, but the phase of the noisy signal is unchanged. This conventional method produces the estimating error because it exploits the noisy phase, especially in low SNR, and it produces “musical noise” because of the inaccuracy of the noise estimation. This paper proposes a multi-band spectral subtraction algorithm based on maximum posteriori phase estimation. Experimental results show that the proposed method can get better performance than the conventional method especially in low SNR.
WANG Jiaching, LIN Changhong, WANG Shufan, et al. Compressive Sensing-based speech enhancement[J]. IEEE Transactions on Audio, Speech and Language Processing, 2016, 24(11): 2122-2131. doi: 10.1109/TASLP.2016.2598306.
[2]
WÓJCICKI K, MILACIC M, STARK A, et al. Exploiting conjugate symmetry of the short-time fourier spectrum for speech enhancement[J]. IEEE Signal Processing Letters, 2008, 15: 461-464. doi: 10.1109/LSP.2008.923579.
[3]
MOWLAEE P and KULMER J. Harmonic phase estimation in single-channel speech enhancement using phase decomposition and SNR information[J]. IEEE Transactions on Audio, Speech and Language Processing, 2015, 23(9): 1521-1532. doi: 10.1109/TASLP.2015.2439038.
[4]
KULMER J and MOWLAEE P. Phase estimation in single channel speech enhancement using phase decomposition[J]. IEEE Signal Processing Letters, 2015, 22(5): 598-602. doi: 10.1109/LSP.2014.2365040.
[5]
BOLLS F. Suppression of acoustic noise in speech using spectral subtraction [J]. IEEE Transactions on Acoustics, Speech and Signal Processing, 1979, 27(2): 113-120. doi: 10.1109/TASSP.1979.1163209.
[6]
WIENER N. The Extrapolation, Interpolation, and Smoothing of Stationary Time Series With Engineering Applications[M]. Cambridge: Massachusetts, MIT, 1949: 81-101.
[7]
EPHRAIM Y and MALAH D. Speech enhancement using a minimum mean-square error short-time spectral amplitude estimator[J]. IEEE Transactions on Acoustics, Speech, and Signal Processing, 1984, 32(6): 1109-1121. doi: 10.1109/ TASSP.1984.1164453.
[8]
KAMATH S and LOIZOU P C. A multi-band spectral subtraction method for enhancing speech corrupted by colored noise[C]. IEEE International Conference on Acoustics, Speech, and Signal Processing, Orlando, FL, USA, 2002: IV-4164-IV-4164.
[9]
VARY P. Noise Suppression by spectral magnitude estimation: Mechanism and theoretical limits[J]. Signal Processing, 1985, 8(4): 387-400. doi: 10.1016/0165-1684(85) 90002-7.
[10]
SAMUI S. Improved single channel phase-aware speech enhancement technique for low signal-to-noise ration signal[J]. IET Signal Processing, 2016, 10(6): 641-650. doi: 10.1049/ iet-spr.2015.0182.
[11]
KULMER J and MOWLAEE P. Harmonic phase estimation in single-channel speech enhancement using Von Mises distribution and prior SNR[C]. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Brisbane, Australia, 2015: 5063-5067.
[12]
KAY S M. Fundamentals of Statistical Signal Processing, Volume I: Estimation Theory[M]. New Jersey: Prentice Hall PTR, 1993: 164-172.
[13]
TAAL C H, HENDRIKS R C, HEUSDENS R, et al. An algorithm for intelligibility prediction of time-frequency weighted noisy speech[J]. IEEE Transactions on Audio, Speech and Languages, 2011, 19(7): 2125-2136. doi: 10.1109/ TASL.2011.2114881.
[14]
GAICH A and MOWLAEE P. On speech quality estimation of phase-aware single-channel speech enhancement[C]. Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing(ICASSP), Brisbane, Australia, 2015: 216-220.
