Eigenvector-Based Speech Mask Estimation for Multi-Channel Speech Enhancement

Pfeifenberger, Lukas; Zöhrer, Matthias; Pernkopf, Franz

doi:10.1109/taslp.2019.2941592

Cited by 23 publications

(25 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While it is possible to select from a broad range of broadband beamformers such as the MVDR or the GSC, we use the GEV for its superior performance in earlier experiments [12]. The GEV beamformer, constrains the filter weights W (k, l) to maximize the SNR ξ(k, l) at the beamformer output [20,21], i.e.…”

Section: Gev Beamformermentioning

confidence: 99%

“…In fact, it can be easily verified that the magnitudes of the BAN and PAN compensation factors are identical. Inserting (12) into (11) gives the GEV-PAN beamformer:…”

Section: Gev Beamformermentioning

confidence: 99%

“…In theory, the compensation factor CP AN turns the GEV into the MVDR beamformer. However, as the former avoids the inversion of ΦNN when using (9), it is numerically more stable and achieves better PESQ and OPS scores [12].…”

Section: Gev Beamformermentioning

confidence: 99%

“…Recent contributions use a spectral gain mask to distinguish between speech and noise signals, which is then used to estimate their respective PSD matrices. Such a speech mask may be obtained using model-based clustering approaches [6][7][8], or data-driven regression [9][10][11][12] based on various types of neural networks (NN). While clustering approaches require some prior knowledge like the array geometry or the statistics of the noise, NNs are able to learn the speech mask from training data, without additional information.…”

Section: Introductionmentioning

confidence: 99%

“…In this paper, we extend our work in [12], where we used several NN architectures to estimate the speech mask using eigenvector features. As the largest NN requires over a million weights to be trained, the aim is to significantly reduce model complexity, while maintaining performance.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Eigenvector-Based Speech Mask Estimation Using Logistic Regression

2017

Self Cite

View full text Add to dashboard Cite

In this paper, we use a logistic regression to learn a speech mask from the dominant eigenvector of the Power Spectral Density (PSD) matrix of a multi-channel speech signal corrupted by ambient noise. We employ this speech mask to construct the Generalized Eigenvalue (GEV) beamformer and a Wiener postfilter. Further, we extend the beamformer to compensate for speech distortions. We do not make any assumptions about the array geometry or the characteristics of the speech and noise sources. Those parameters are learned from training data. Our assumptions are that the speaker may move slowly in the nearfield of the array, and that the noise is in the far-field. We compare our speech enhancement system against recent contributions using the CHiME4 corpus. We show that our approach yields superior results, both in terms of perceptual speech quality and speech mask estimation error.

show abstract

Section: Gev Beamformermentioning

confidence: 99%

“…In fact, it can be easily verified that the magnitudes of the BAN and PAN compensation factors are identical. Inserting (12) into (11) gives the GEV-PAN beamformer:…”

Section: Gev Beamformermentioning

confidence: 99%

Section: Gev Beamformermentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Eigenvector-Based Speech Mask Estimation Using Logistic Regression

2017

Self Cite

View full text Add to dashboard Cite

show abstract

Dual-Channel Speech Enhancement Using Neural Network Adaptive Beamforming

Jiang

Liu

Shuai

et al. 2022

Communications and Networking

View full text Add to dashboard Cite

Dual-channel speech enhancement based on traditional beamforming is difficult to effectively suppress noise. In recent years, it is promising to replace beamforming with a neural network that learns spectral characteristic. This paper proposes a neural network adaptive beamforming end-to-end dual-channel model for speech enhancement task. First, the LSTM layer is used to directly process the original speech waveform to estimate the time-domain beamforming filter coefficients of each channel and convolve and sum it with the input speech. Second, we modified a fully-convolutional time-domain audio separation network (Conv-TasNet) into a network suitable for speech enhancement which is called Denoising-TasNet to further enhance the output of the beamforming. The experimental results show that the proposed method is better than convolutional recurrent network (CRN) model and several popular noise reduction methods.

show abstract

Speech enhancement using long short term memory with trained speech features and adaptive wiener filter

Garg

2022

Multimed Tools Appl

View full text Add to dashboard Cite

Speech enhancement is the process of enhancing the clarity and intelligibility of speech signals that have been degraded due to background noise. With the assistance of deep learning, a novel speech signal enhancement model is introduced in this research. The proposed model is divided into two phases: (i) Training (ii) Testing. In the training phase, the noise spectrum and signal spectrum are estimated via a Non-negative Matrix Factorization (NMF) from the noisy input signal. Then, Empirical Mean Decomposition (EMD) features are extracted from the Wiener filter. The de-noised signal is acquired from EMD, the bark frequency is evaluated and the Fractional Delta AMS features are extracted. The key contribution of this study is the use of the Long Short Term Memory (LSTM) model to properly estimate the tuning factor η of the Wiener filter for all input signals. The LSTM was trained by the extracted features (EMD) via a modified wiener filter for decomposing the spectral signal and the output of EMD is the denoised enhanced speech signal. A comparative evaluation is carried out between the proposed and existing models in terms of error measures.

show abstract

Eigenvector-Based Speech Mask Estimation for Multi-Channel Speech Enhancement

Cited by 23 publications

References 52 publications

Eigenvector-Based Speech Mask Estimation Using Logistic Regression

Eigenvector-Based Speech Mask Estimation Using Logistic Regression

Dual-Channel Speech Enhancement Using Neural Network Adaptive Beamforming

Speech enhancement using long short term memory with trained speech features and adaptive wiener filter

Contact Info

Product

Resources

About