Issues in Frequency Domain Blind Source Separation á- A Critical Revisit

Robledo-Arnuncio, Enrique; Juang, Biing‐Hwang

doi:10.1109/icassp.2005.1416295

Cited by 4 publications

(5 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In practice, the separation performance is mainly affected by the global convergence. Nevertheless, the influence of the causality on experimental results has been pointed out [3,9]. It is important to keep in mind the physical interpretation of the stability conditions (28) and (30).…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

An LMS viewpoint on the local stability of second order blind source separation

Bourgeois

2005

IEEE/SP 13th Workshop on Statistical Signal Processing, 2005

View full text Add to dashboard Cite

The local stability of blind source separation (BSS) algorithms has been mostly investigated in the vicinity of the mixing system inverse, i.e. assuming not only separation but also deconvolution of the sources. In the case of convolutive mixtures, other equilibrium points exist that do not deconvolve the sources. In this paper, we examine the local stability at such an equilibrium point for a second-order BSS algorithm based on the source non-stationarity. Considering explicitly source silences (for example with speech sources) greatly simplifies the obtained local stability conditions and allows a comparison with the classical LMS noise canceller. This analysis reveals a discrepancy between causal and double-sided (acausal) mixing channels.

show abstract

Section: Discussionmentioning

confidence: 99%

“…Such a formulation of the mixing system hides its causality. Nevertheless, the performance of frequency-domain BSS algorithm is also affected by the causality of the mixing/separation system [9].…”

Section: Discussionmentioning

confidence: 99%

An LMS viewpoint on the local stability of second order blind source separation

Bourgeois

2005

IEEE/SP 13th Workshop on Statistical Signal Processing, 2005

View full text Add to dashboard Cite

show abstract

“…A method that avoids the circularity effects but maintains the computational efficiency of the FFT has been presented in [262]. Further discussion on the circularity problem can be found in [189].…”

Section: Circularity Problemmentioning

confidence: 99%

Convolutive Blind Source Separation Methods

Pedersen

Larsen

Kjems

et al. 2008

Springer Handbook of Speech Processing

138

129

View full text Add to dashboard Cite

show abstract

“…The circular convolution describes the linear convolution accurately after zero-padding [133]. Without zero-padding, there is a discrepancy in the approximation of a linear convolution by using a circular convolution due to sampling the continuous frequency representation of a finite length signal [152]. This artifact is known as a circularity effect.…”

Section: The Hfrtf-domain Representationmentioning

confidence: 99%

Low Latency Convolutive Blind Source Separation

Chua¹

View full text Add to dashboard Cite

<p>In most real-time systems, particularly for applications involving system identification, latency is a critical issue. These applications include, but are not limited to, blind source separation (BSS), beamforming, speech dereverberation, acoustic echo cancellation and channel equalization. The system latency consists of an algorithmic delay and an estimation computational time. The latter can be avoided by using a multi-thread system, which runs the estimation process and the processing procedure simultaneously. The former, which consists of a delay of one window length, is usually unavoidable for the frequency-domain approaches. For frequency-domain approaches, a block of data is acquired by using a window, transformed and processed in the frequency domain, and recovered back to the time domain by using an overlap-add technique. In the frequency domain, the convolutive model, which is usually used to describe the process of a linear time-invariant (LTI) system, can be represented by a series of multiplicative models to facilitate estimation. To implement frequency-domain approaches in real-time applications, the short-time Fourier transform (STFT) is commonly used. The window used in the STFT must be at least twice the room impulse response which is long, so that the multiplicative model is sufficiently accurate. The delay constraint caused by the associated blockwise processing window length makes most the frequency-domain approaches inapplicable for real-time systems. This thesis aims to design a BSS system that can be used in a real-time scenario with minimal latency. Existing BSS approaches can be integrated into our system to perform source separation with low delay without affecting the separation performance. The second goal is to design a BSS system that can perform source separation in a non-stationary environment. We first introduce a subspace approach to directly estimate the separation parameters in the low-frequency-resolution time-frequency (LFRTF) domain. In the LFRTF domain, a shorter window is used to reduce the algorithmic delay of the system during the signal acquisition, e.g., the window length is shorter than the room impulse response. The subspace method facilitates the deconvolution of a convolutive mixture to a new instantaneous mixture and simplifies the estimation process. Second, we propose an alternative approach to address the algorithmic latency problem. The alternative method enables us to obtain the separation parameters in the LFRTF domain based on parameters estimated in the high-frequency-resolution time-frequency (HFRTF) domain, where the window length is longer than the room impulse response, without affecting the separation performance. The thesis also provides a solution to address the BSS problem in a non-stationary environment. We utilize the ``meta-information" that is obtained from previous BSS operations to facilitate the separation in the future without performing the entire BSS process again. Repeating a BSS process can be computationally expensive. Most conventional BSS algorithms require sufficient signal samples to perform analysis and this prolongs the estimation delay. By utilizing information from the entire spectrum, our method enables us to update the separation parameters with only a single snapshot of observation data. Hence, our method minimizes the estimation period, reduces the redundancy and improves the efficacy of the system. The final contribution of the thesis is a non-iterative method for impulse response shortening. This method allows us to use a shorter representation to approximate the long impulse response. It further improves the computational efficiency of the algorithm and yet achieves satisfactory performance.</p>

show abstract

Issues in Frequency Domain Blind Source Separation á- A Critical Revisit

Cited by 4 publications

References 4 publications

An LMS viewpoint on the local stability of second order blind source separation

An LMS viewpoint on the local stability of second order blind source separation

Convolutive Blind Source Separation Methods

Low Latency Convolutive Blind Source Separation

Contact Info

Product

Resources

About