Distant-Talking Speech Recognition Based on Spectral Subtraction by Multi-Channel LMS Algorithm

Abstract

SUMMARYWe propose a blind dereverberation method based on spectral subtraction using a multi-channel least mean squares (MCLMS) algorithm for distant-talking speech recognition. In a distant-talking environment, the channel impulse response is longer than the short-term spectral analysis window. By treating the late reverberation as additive noise, a noise reduction technique based on spectral subtraction was proposed to estimate the power spectrum of the clean speech using power spectra of the distorted speech and the unknown impulse responses. To estimate the power spectra of the impulse responses, a variable step-size unconstrained MCLMS (VSS-UMCLMS) algorithm for identifying the impulse responses in a time domain is extended to a frequency domain. To reduce the effect of the estimation error of the channel impulse response, we normalize the early reverberation by cepstral mean normalization (CMN) instead of spectral subtraction using the estimated impulse response. Furthermore, our proposed method is combined with conventional delay-andsum beamforming. We conducted recognition experiments on a distorted speech signal simulated by convolving multi-channel impulse responses with clean speech. The proposed method achieved a relative error reduction rate of 22.4% in relation to conventional CMN. By combining the proposed method with beamforming, a relative error reduction rate of 24.5% in relation to the conventional CMN with beamforming was achieved using only an isolated word (with duration of about 0.6 s) to estimate the spectrum of the impulse response.

Deep neural network-based bottleneck feature and denoising autoencoder-based dereverberation for distant-talking speaker identification

“…For method 1, only CMN with beamforming was used to reduce the reverberation (denoted as 'CMN'). For comparison, MCLMS-SS- [32] and MSLP-SS [17]-based dereverberation was performed in method 2 and method 3, respectively. The MCLMS-SS and MSLP-SS methods both treated late reverberation as additional noise and used the spectral subtraction method to suppress it.…”

“…This method first estimates late reverberation using longterm MSLP and then suppresses this with the subsequent spectral subtraction. Wang et al proposed a distanttalking speech recognition method based on generalized spectral subtraction (SS) [30] employing the multichannel least mean squares (MCLMS) algorithm [13,31,32]. The authors further extended their method to distanttalking speaker recognition and proposed an efficient computational method for combining the likelihoods of dereverberant speech using multiple compensation parameter sets [23].…”

Application of an Improved GSC in Tte Communication With Antenna Array

“…Keep T1 stationary but move T2 to change the parameter ϕ. To make a wide comparison with other techniques, this paper adds multi-channel LMS and RLS algorithms as comparison groups [17,18].…”

Environment-dependent denoising autoencoder for distant-talking speech recognition

“…There are two different approaches, namely, front-and back-end-based methods, for dealing with this problem [1]. Many front-end-based approaches [1][2][3][4][5][6][7][8] have been proposed to reduce the effect of reverberation *Correspondence: wang@vos.nagaokaut.ac.jp 2 Nagaoka University of Technology, 1603-1 Kamitomioka, Nagaoka 940-2188, Japan Full list of author information is available at the end of the article in the observed speech signal. The back-end-based methods, on the other hand, attempt to modify the acoustic model and/or decoder to suit the respective reverberant environment [9,10].…”

“…Many single-and multi-channel dereverberation methods have been proposed to suppress reverberation [2][3][4][11][12][13][14]. Single-channel dereverberation approaches are much easier and cheaper to implement in real applications than multi-channel ones.…”