Sorry, you need to enable JavaScript to visit this website.

Source Separation and Signal Enhancement

Speech Enhancement with Binaural Cues Derived from a Priori Codebook


In conventional codebook-driven speech enhancement, only spectral envelopes of speech and noise are considered, and at the same time, the type of noise is the priori information when we enhance the noisy speech. In this paper, we propose a novel codebook-based speech enhancement method which exploits a priori information about binaural cues, including clean cue and pre-enhanced cue, stored in the trained codebook. This method includes two main parts: offline training of cues and online enhancement by means of cues.

Paper Details

Authors:
Nan Chen,Changchun Bao, Feng Deng
Submitted On:
13 October 2016 - 9:25pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

ISLSLP2016 陈楠.ppt

(0)

Keywords

Subscribe

[1] Nan Chen,Changchun Bao, Feng Deng, "Speech Enhancement with Binaural Cues Derived from a Priori Codebook", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1152. Accessed: Sep. 25, 2017.
@article{1152-16,
url = {http://sigport.org/1152},
author = {Nan Chen;Changchun Bao; Feng Deng },
publisher = {IEEE SigPort},
title = {Speech Enhancement with Binaural Cues Derived from a Priori Codebook},
year = {2016} }
TY - EJOUR
T1 - Speech Enhancement with Binaural Cues Derived from a Priori Codebook
AU - Nan Chen;Changchun Bao; Feng Deng
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1152
ER -
Nan Chen,Changchun Bao, Feng Deng. (2016). Speech Enhancement with Binaural Cues Derived from a Priori Codebook. IEEE SigPort. http://sigport.org/1152
Nan Chen,Changchun Bao, Feng Deng, 2016. Speech Enhancement with Binaural Cues Derived from a Priori Codebook. Available at: http://sigport.org/1152.
Nan Chen,Changchun Bao, Feng Deng. (2016). "Speech Enhancement with Binaural Cues Derived from a Priori Codebook." Web.
1. Nan Chen,Changchun Bao, Feng Deng. Speech Enhancement with Binaural Cues Derived from a Priori Codebook [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1152

A source/filter model with adaptive constraints for NMF-based speech separation [slides]

Paper Details

Authors:
Damien Bouvier, Nicolas Obin, Axel Roebel, Marco Liuni
Submitted On:
29 March 2016 - 10:22am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

ICASSP16_3106_slides.pdf

(191 downloads)

Keywords

Subscribe

[1] Damien Bouvier, Nicolas Obin, Axel Roebel, Marco Liuni, "A source/filter model with adaptive constraints for NMF-based speech separation [slides]", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1067. Accessed: Sep. 25, 2017.
@article{1067-16,
url = {http://sigport.org/1067},
author = {Damien Bouvier; Nicolas Obin; Axel Roebel; Marco Liuni },
publisher = {IEEE SigPort},
title = {A source/filter model with adaptive constraints for NMF-based speech separation [slides]},
year = {2016} }
TY - EJOUR
T1 - A source/filter model with adaptive constraints for NMF-based speech separation [slides]
AU - Damien Bouvier; Nicolas Obin; Axel Roebel; Marco Liuni
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1067
ER -
Damien Bouvier, Nicolas Obin, Axel Roebel, Marco Liuni. (2016). A source/filter model with adaptive constraints for NMF-based speech separation [slides]. IEEE SigPort. http://sigport.org/1067
Damien Bouvier, Nicolas Obin, Axel Roebel, Marco Liuni, 2016. A source/filter model with adaptive constraints for NMF-based speech separation [slides]. Available at: http://sigport.org/1067.
Damien Bouvier, Nicolas Obin, Axel Roebel, Marco Liuni. (2016). "A source/filter model with adaptive constraints for NMF-based speech separation [slides]." Web.
1. Damien Bouvier, Nicolas Obin, Axel Roebel, Marco Liuni. A source/filter model with adaptive constraints for NMF-based speech separation [slides] [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1067

Deep Unfolding for Multichannel Source Separation


Title slide for Deep Unfolding for Multichannel Source Separation

Deep unfolding has recently been proposed to derive novel deep network architectures from model-based approaches. In this paper, we consider its application to multichannel source separation. We unfold a multichannel Gaussian mixture model (MCGMM), resulting in a deep MCGMM computational network that directly processes complex-valued frequency-domain multichannel audio and has an architecture defined explicitly by a generative model, thus combining the advantages of deep networks and model-based approaches.

Paper Details

Authors:
Scott Wisdom, John R. Hershey, Jonathan Le Roux, Shinji Watanabe
Submitted On:
24 March 2016 - 9:11pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

WisdomHersheyLeRouxWatanabe_ICASSP2016_publish.pdf

(199 downloads)

Keywords

Subscribe

[1] Scott Wisdom, John R. Hershey, Jonathan Le Roux, Shinji Watanabe, "Deep Unfolding for Multichannel Source Separation", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1031. Accessed: Sep. 25, 2017.
@article{1031-16,
url = {http://sigport.org/1031},
author = {Scott Wisdom; John R. Hershey; Jonathan Le Roux; Shinji Watanabe },
publisher = {IEEE SigPort},
title = {Deep Unfolding for Multichannel Source Separation},
year = {2016} }
TY - EJOUR
T1 - Deep Unfolding for Multichannel Source Separation
AU - Scott Wisdom; John R. Hershey; Jonathan Le Roux; Shinji Watanabe
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1031
ER -
Scott Wisdom, John R. Hershey, Jonathan Le Roux, Shinji Watanabe. (2016). Deep Unfolding for Multichannel Source Separation. IEEE SigPort. http://sigport.org/1031
Scott Wisdom, John R. Hershey, Jonathan Le Roux, Shinji Watanabe, 2016. Deep Unfolding for Multichannel Source Separation. Available at: http://sigport.org/1031.
Scott Wisdom, John R. Hershey, Jonathan Le Roux, Shinji Watanabe. (2016). "Deep Unfolding for Multichannel Source Separation." Web.
1. Scott Wisdom, John R. Hershey, Jonathan Le Roux, Shinji Watanabe. Deep Unfolding for Multichannel Source Separation [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1031

JOINTLY OPTIMAL NEAR-END AND FAR-END MULTI-MICROPHONE SPEECH INTELLIGIBILITY ENHANCEMENT BASED ON MUTUAL INFORMATION

Paper Details

Authors:
Richard C. Hendriks, W. Bastiaan Kle
Submitted On:
23 March 2016 - 7:24pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

ICASSP16_Poster_Seyran.pdf

(163 downloads)

Keywords

Subscribe

[1] Richard C. Hendriks, W. Bastiaan Kle, "JOINTLY OPTIMAL NEAR-END AND FAR-END MULTI-MICROPHONE SPEECH INTELLIGIBILITY ENHANCEMENT BASED ON MUTUAL INFORMATION", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1006. Accessed: Sep. 25, 2017.
@article{1006-16,
url = {http://sigport.org/1006},
author = {Richard C. Hendriks; W. Bastiaan Kle },
publisher = {IEEE SigPort},
title = {JOINTLY OPTIMAL NEAR-END AND FAR-END MULTI-MICROPHONE SPEECH INTELLIGIBILITY ENHANCEMENT BASED ON MUTUAL INFORMATION},
year = {2016} }
TY - EJOUR
T1 - JOINTLY OPTIMAL NEAR-END AND FAR-END MULTI-MICROPHONE SPEECH INTELLIGIBILITY ENHANCEMENT BASED ON MUTUAL INFORMATION
AU - Richard C. Hendriks; W. Bastiaan Kle
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1006
ER -
Richard C. Hendriks, W. Bastiaan Kle. (2016). JOINTLY OPTIMAL NEAR-END AND FAR-END MULTI-MICROPHONE SPEECH INTELLIGIBILITY ENHANCEMENT BASED ON MUTUAL INFORMATION. IEEE SigPort. http://sigport.org/1006
Richard C. Hendriks, W. Bastiaan Kle, 2016. JOINTLY OPTIMAL NEAR-END AND FAR-END MULTI-MICROPHONE SPEECH INTELLIGIBILITY ENHANCEMENT BASED ON MUTUAL INFORMATION. Available at: http://sigport.org/1006.
Richard C. Hendriks, W. Bastiaan Kle. (2016). "JOINTLY OPTIMAL NEAR-END AND FAR-END MULTI-MICROPHONE SPEECH INTELLIGIBILITY ENHANCEMENT BASED ON MUTUAL INFORMATION." Web.
1. Richard C. Hendriks, W. Bastiaan Kle. JOINTLY OPTIMAL NEAR-END AND FAR-END MULTI-MICROPHONE SPEECH INTELLIGIBILITY ENHANCEMENT BASED ON MUTUAL INFORMATION [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1006

An Expectation-Maximization Eigenvector Clustering Approach to Direction of Arrival Estimation of Multiple Speech Sources

Paper Details

Authors:
Xiong Xiao, Shengkui Zhao, Thi Ngoc Tho Nguyen, Douglas L. Jones, Eng Siong Chng, Haizhou Li
Submitted On:
22 March 2016 - 2:37am
Short Link:
Type:
Event:
Presenter's Name:
Document Year:
Cite

Document Files

ICASSP16_multiDOA.pdf

(161 downloads)

Keywords

Subscribe

[1] Xiong Xiao, Shengkui Zhao, Thi Ngoc Tho Nguyen, Douglas L. Jones, Eng Siong Chng, Haizhou Li, "An Expectation-Maximization Eigenvector Clustering Approach to Direction of Arrival Estimation of Multiple Speech Sources", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/956. Accessed: Sep. 25, 2017.
@article{956-16,
url = {http://sigport.org/956},
author = {Xiong Xiao; Shengkui Zhao; Thi Ngoc Tho Nguyen; Douglas L. Jones; Eng Siong Chng; Haizhou Li },
publisher = {IEEE SigPort},
title = {An Expectation-Maximization Eigenvector Clustering Approach to Direction of Arrival Estimation of Multiple Speech Sources},
year = {2016} }
TY - EJOUR
T1 - An Expectation-Maximization Eigenvector Clustering Approach to Direction of Arrival Estimation of Multiple Speech Sources
AU - Xiong Xiao; Shengkui Zhao; Thi Ngoc Tho Nguyen; Douglas L. Jones; Eng Siong Chng; Haizhou Li
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/956
ER -
Xiong Xiao, Shengkui Zhao, Thi Ngoc Tho Nguyen, Douglas L. Jones, Eng Siong Chng, Haizhou Li. (2016). An Expectation-Maximization Eigenvector Clustering Approach to Direction of Arrival Estimation of Multiple Speech Sources. IEEE SigPort. http://sigport.org/956
Xiong Xiao, Shengkui Zhao, Thi Ngoc Tho Nguyen, Douglas L. Jones, Eng Siong Chng, Haizhou Li, 2016. An Expectation-Maximization Eigenvector Clustering Approach to Direction of Arrival Estimation of Multiple Speech Sources. Available at: http://sigport.org/956.
Xiong Xiao, Shengkui Zhao, Thi Ngoc Tho Nguyen, Douglas L. Jones, Eng Siong Chng, Haizhou Li. (2016). "An Expectation-Maximization Eigenvector Clustering Approach to Direction of Arrival Estimation of Multiple Speech Sources." Web.
1. Xiong Xiao, Shengkui Zhao, Thi Ngoc Tho Nguyen, Douglas L. Jones, Eng Siong Chng, Haizhou Li. An Expectation-Maximization Eigenvector Clustering Approach to Direction of Arrival Estimation of Multiple Speech Sources [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/956

Blind Speech Separation 
based on Complex Spherical k-Mode Clustering


We present an algorithm for clustering complex-valued unit length vectors on the unit hypersphere, which we call complex spherical k-mode clustering, as it can be viewed as a generalization of the spherical k-means algorithm to normalized complex-valued vectors. We show how the proposed algorithm can be derived from the Expectation Maximization algorithm for complex Watson mixture models and prove its applicability in a blind speech separation (BSS) task with real-world room impulse response measurements.

Paper Details

Authors:
Lukas Drude, Christoph Boeddeker, Reinhold Haeb-Umbach
Submitted On:
20 March 2016 - 5:37am
Short Link:
Type:
Event:
Presenter's Name:
Document Year:
Cite

Document Files

2016-03-15_icassp_bss.pdf

(180 downloads)

Keywords

Subscribe

[1] Lukas Drude, Christoph Boeddeker, Reinhold Haeb-Umbach, "Blind Speech Separation 
based on Complex Spherical k-Mode Clustering", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/864. Accessed: Sep. 25, 2017.
@article{864-16,
url = {http://sigport.org/864},
author = {Lukas Drude; Christoph Boeddeker; Reinhold Haeb-Umbach },
publisher = {IEEE SigPort},
title = {Blind Speech Separation 
based on Complex Spherical k-Mode Clustering},
year = {2016} }
TY - EJOUR
T1 - Blind Speech Separation 
based on Complex Spherical k-Mode Clustering
AU - Lukas Drude; Christoph Boeddeker; Reinhold Haeb-Umbach
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/864
ER -
Lukas Drude, Christoph Boeddeker, Reinhold Haeb-Umbach. (2016). Blind Speech Separation 
based on Complex Spherical k-Mode Clustering. IEEE SigPort. http://sigport.org/864
Lukas Drude, Christoph Boeddeker, Reinhold Haeb-Umbach, 2016. Blind Speech Separation 
based on Complex Spherical k-Mode Clustering. Available at: http://sigport.org/864.
Lukas Drude, Christoph Boeddeker, Reinhold Haeb-Umbach. (2016). "Blind Speech Separation 
based on Complex Spherical k-Mode Clustering." Web.
1. Lukas Drude, Christoph Boeddeker, Reinhold Haeb-Umbach. Blind Speech Separation 
based on Complex Spherical k-Mode Clustering [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/864

Neural Network based Spectral Mask Estimation for Acoustic Beamforming


We present a neural network based approach to acoustic beamform- ing. The network is used to estimate spectral masks from which the Cross-Power Spectral Density matrices of speech and noise are estimated, which in turn are used to compute the beamformer co- efficients. The network training is independent of the number and the geometric configuration of the microphones. We further show that it is possible to train the network on clean speech only, avoid- ing the need for stereo data with separated speech and noise. Two types of networks are evaluated.

Paper Details

Authors:
Reinhold Haeb-Umbach
Submitted On:
20 March 2016 - 5:37am
Short Link:
Type:
Event:
Presenter's Name:
Document Year:
Cite

Document Files

icassp_2016.pdf

(172 downloads)

Keywords

Subscribe

[1] Reinhold Haeb-Umbach, "Neural Network based Spectral Mask Estimation for Acoustic Beamforming", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/863. Accessed: Sep. 25, 2017.
@article{863-16,
url = {http://sigport.org/863},
author = {Reinhold Haeb-Umbach },
publisher = {IEEE SigPort},
title = {Neural Network based Spectral Mask Estimation for Acoustic Beamforming},
year = {2016} }
TY - EJOUR
T1 - Neural Network based Spectral Mask Estimation for Acoustic Beamforming
AU - Reinhold Haeb-Umbach
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/863
ER -
Reinhold Haeb-Umbach. (2016). Neural Network based Spectral Mask Estimation for Acoustic Beamforming. IEEE SigPort. http://sigport.org/863
Reinhold Haeb-Umbach, 2016. Neural Network based Spectral Mask Estimation for Acoustic Beamforming. Available at: http://sigport.org/863.
Reinhold Haeb-Umbach. (2016). "Neural Network based Spectral Mask Estimation for Acoustic Beamforming." Web.
1. Reinhold Haeb-Umbach. Neural Network based Spectral Mask Estimation for Acoustic Beamforming [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/863

NMF-based source separation utilizing prior knowledge on encoding vector

Paper Details

Authors:
Submitted On:
19 March 2016 - 4:40am
Short Link:
Type:

Document Files

ICASSP2016_포스터_권기수_pdf.pdf

(0)

Keywords

Subscribe

[1] , "NMF-based source separation utilizing prior knowledge on encoding vector", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/790. Accessed: Sep. 25, 2017.
@article{790-16,
url = {http://sigport.org/790},
author = { },
publisher = {IEEE SigPort},
title = {NMF-based source separation utilizing prior knowledge on encoding vector},
year = {2016} }
TY - EJOUR
T1 - NMF-based source separation utilizing prior knowledge on encoding vector
AU -
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/790
ER -
. (2016). NMF-based source separation utilizing prior knowledge on encoding vector. IEEE SigPort. http://sigport.org/790
, 2016. NMF-based source separation utilizing prior knowledge on encoding vector. Available at: http://sigport.org/790.
. (2016). "NMF-based source separation utilizing prior knowledge on encoding vector." Web.
1. . NMF-based source separation utilizing prior knowledge on encoding vector [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/790

Variable Span Filtering for Speech Enhancement


In this work, we consider enhancement of multichannel speech recordings. Linear filtering and subspace approaches have been considered previously for solving the problem. The current linear filtering methods, although many variants exist, have limited control of noise reduction and speech distortion. Subspace approaches, on the other hand, can potentially yield better control by filtering in the eigen-domain, but traditionally these approaches have not been optimized explicitly for traditional noise reduction and signal distortion measures.

Paper Details

Authors:
Jacob Benesty, Mads Græsbøll Christensen
Submitted On:
18 March 2016 - 10:24am
Short Link:
Type:
Event:
Presenter's Name:
Document Year:
Cite

Document Files

icassp2016varSpan_jrj.pdf

(155 downloads)

Keywords

Subscribe

[1] Jacob Benesty, Mads Græsbøll Christensen, "Variable Span Filtering for Speech Enhancement", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/764. Accessed: Sep. 25, 2017.
@article{764-16,
url = {http://sigport.org/764},
author = {Jacob Benesty; Mads Græsbøll Christensen },
publisher = {IEEE SigPort},
title = {Variable Span Filtering for Speech Enhancement},
year = {2016} }
TY - EJOUR
T1 - Variable Span Filtering for Speech Enhancement
AU - Jacob Benesty; Mads Græsbøll Christensen
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/764
ER -
Jacob Benesty, Mads Græsbøll Christensen. (2016). Variable Span Filtering for Speech Enhancement. IEEE SigPort. http://sigport.org/764
Jacob Benesty, Mads Græsbøll Christensen, 2016. Variable Span Filtering for Speech Enhancement. Available at: http://sigport.org/764.
Jacob Benesty, Mads Græsbøll Christensen. (2016). "Variable Span Filtering for Speech Enhancement." Web.
1. Jacob Benesty, Mads Græsbøll Christensen. Variable Span Filtering for Speech Enhancement [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/764

NMF-based source separation utilizing prior knowledge on encoding vector

Paper Details

Authors:
Submitted On:
18 March 2016 - 7:30am
Short Link:
Type:
Event:
Presenter's Name:
Document Year:
Cite

Document Files

ICASSP2016_포스터_권기수pptx.pptx

(0)

Keywords

Subscribe

[1] , "NMF-based source separation utilizing prior knowledge on encoding vector", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/762. Accessed: Sep. 25, 2017.
@article{762-16,
url = {http://sigport.org/762},
author = { },
publisher = {IEEE SigPort},
title = {NMF-based source separation utilizing prior knowledge on encoding vector},
year = {2016} }
TY - EJOUR
T1 - NMF-based source separation utilizing prior knowledge on encoding vector
AU -
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/762
ER -
. (2016). NMF-based source separation utilizing prior knowledge on encoding vector. IEEE SigPort. http://sigport.org/762
, 2016. NMF-based source separation utilizing prior knowledge on encoding vector. Available at: http://sigport.org/762.
. (2016). "NMF-based source separation utilizing prior knowledge on encoding vector." Web.
1. . NMF-based source separation utilizing prior knowledge on encoding vector [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/762

Pages