Rank-1 constrained Multichannel Wiener Filter for speech recognition in noisy environments. (May 2018)

• Record Type: Journal Article
• Authors: Wang, Ziteng; Vincent, Emmanuel; Serizel, Romain; Yan, Yonghong
• Is Part Of: Computer speech & language. Volume 49(2018)
• Page Start: 37
• Page End: 51
• Publication Date: 2018-05
• Subjects: Rank-1 Multichannel Wiener Filter -- Speech recognition -- Residual noise power -- Deep neural network; Speech processing systems -- Periodicals; Automatic speech recognition -- Periodicals; Computers -- Periodicals; Linguistics -- Periodicals; Speech-Language Pathology -- Periodicals; Traitement automatique de la parole -- Périodiques; Reconnaissance automatique de la parole -- Périodiques; Automatic speech recognition; Speech processing systems; Electronic journals; Periodicals; 006.454
• DOI: 10.1016/j.csl.2017.11.003
• Languages: English
• ISSNs: 0885-2308

Abstract

Highlights: A constant residual noise power constraint for rank-1 MWF is proposed. Speech covariance matrix reconstruction to fulfill the rank-1 assumption. An extensive comparison of the multichannel linear filters supported by BLSTM. A feature variance metric that correlates with the word error rate. Abstract: Multichannel linear filters, such as the Multichannel Wiener Filter (MWF) and the Generalized Eigenvalue (GEV) beamformer are popular signal processing techniques which can improve speech recognition performance. In this paper, we present an experimental study on these linear filters in a specific speech recognition task, namely the CHiME-4 challenge, which features real recordings in multiple noisy environments. Specifically, the rank-1 MWF is employed for noise reduction and a new constant residual noise power constraint is derived which enhances the recognition performance. To fulfill the underlying rank-1 assumption, the speech covariance matrix is reconstructed based on eigenvectors or generalized eigenvectors. Then the rank-1 constrained MWF is evaluated with alternative multichannel linear filters under the same framework, which involves a Bidirectional Long Short-Term Memory (BLSTM) network for mask estimation. The proposed filter outperforms alternative ones, leading to a 40% relative Word Error Rate (WER) reduction compared with the baseline Weighted Delay and Sum (WDAS) beamformer on the real test set, and a 15% relative WER reduction compared with the GEV-BAN method. The results also suggest that the speech recognition accuracy correlates more with the Mel-frequency cepstral coefficients (MFCC) feature variance than with the noise reduction or the speech distortion level.