• Corpus ID: 215416338

MM Algorithms for Joint Independent Subspace Analysis with Application to Blind Single and Multi-Source Extraction

@article{Scheibler2020MMAF,
  title={MM Algorithms for Joint Independent Subspace Analysis with Application to Blind Single and Multi-Source Extraction},
  author={Robin Scheibler and Nobutaka Ono},
  journal={ArXiv},
  year={2020},
  volume={abs/2004.03926}
}
In this work, we propose efficient algorithms for joint independent subspace analysis (JISA), an extension of independent component analysis that deals with parallel mixtures, where not all the components are independent. We derive an algorithmic framework for JISA based on the majorization-minimization (MM) optimization technique (JISA-MM). We use a well-known inequality for super-Gaussian sources to derive a surrogate function of the negative log-likelihood of the observed data. The… 

Figures and Tables from this paper

Independent Vector Analysis via Log-Quadratically Penalized Quadratic Minimization

This work introduces iterative projection with adjustment (IPA), where one demixing filter is updated and all the other sources are adjusted along its current direction, and efficiently decreases the value of the surrogate function.

Independent Vector Extraction for Fast Joint Blind Source Separation and Dereverberation

It is shown that, given the power spectrum for each source, the optimization problem of the new method can be reduced to that of IVE by exploiting the stationary condition, which makes the optimization easy to handle and computationally efficient.

ISS2: An Extension of Iterative Source Steering Algorithm for Majorization-Minimization-Based Independent Vector Analysis

Numerical experiments to separate reverberant speech mixtures show that the ISS 2 converges in fewer MM iterations than the conventional ISS, and is comparable to IP 2.

End-to-End Multi-speaker ASR with Independent Vector Analysis

An end-to-end system for multi-channel, multi-speaker automatic speech recognition that uses the fast and stable iterative source steering algorithm together with a neural source model and demonstrates competitive performance with previous systems using neural beamforming frontends.

Efficient Independent Vector Extraction of Dominant Target Speech

This paper proposes a computationally efficient blind speech extraction method based on a proper modification of the commonly utilized independent vector analysis algorithm, under the mild assumption that the average power of signal of interest outweighs interfering speech sources.

Block Coordinate Descent Algorithms for Auxiliary-Function-Based Independent Vector Extraction

This paper improves the conventional BCDs for IVE by carefully exploiting the stationarity of the Gaussian noise components and develops a BCD for a semiblind IVE in which the transfer functions for several super-Gaussian sources are given a priori.

Over-determined Speech Source Separation and Dereverberation

  • M. TogamiRobin Scheibler
  • Physics
    2020 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC)
  • 2020
It is revealed that an orthogonal constraint enables efficient update of a noise reduction filter in the proposed framework similar to the previously proposed over-determined speech source separation case.

Over-Determined Semi-Blind Speech Source Separation

  • M. TogamiRobin Scheibler
  • Physics
    2021 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC)
  • 2021
We propose a semi-blind speech source separation that jointly optimizes several acoustic functions, i.e., speech source separation (SS), dereverberation (DR), acoustic echo reduction (AE), and

References

SHOWING 1-10 OF 61 REFERENCES

Joint Independent Subspace Analysis Using Second-Order Statistics

  • D. LahatC. Jutten
  • Computer Science, Mathematics
    IEEE Transactions on Signal Processing
  • 2016
This paper combines IVA with ISA and term this new model joint independent subspace analysis (JISA), which provides full performance analysis of JISA, including closed-form expressions for minimal mean square error (MSE), Fisher information and Cramér-Rao lower bound, in the separation of Gaussian data.

A Unified Probabilistic View on Spatially Informed Source Separation and Extraction Based on Independent Vector Analysis

This paper addresses the problem of blind source separation by incorporating prior knowledge into the adaptation of the demixing filters, e.g., the position of the sources, in a probabilistic framework based on Independent Vector Analysis (IVA) as it elegantly and successfully deals with the internal permutation problem.

A Unified Bayesian View on Spatially Informed Source Separation and Extraction based on Independent Vector Analysis

This paper addresses the problem of Blind Source Separation by incorporating prior knowledge into the adaptation of the demixing filters, e.g., the position of the sources, in a Bayesian framework as it elegantly and successfully deals with the internal permutation problem.

Overdetermined Independent Vector Analysis

We address the convolutive blind source separation problem for the (over-)determined case where (i) the number of nonstationary target-sources K is less than that of microphones M, and (ii) there are

Fast Algorithm for Blind Independence-Based Extraction of a Moving Speaker

The experiments corroborate the applicability of the CSV mixing model for the blind moving source extraction as well as the improved convergence of the proposed algorithms.

Blind Source Separation Exploiting Higher-Order Frequency Dependencies

A new algorithm is proposed that exploits higher order frequency dependencies of source signals in order to separate them when they are mixed and outperforms the others in most cases.

Joint Independent Subspace Analysis: A Quasi-Newton Algorithm

A quasi-Newton QN algorithm for joint independent subspace analysis JISA, a recently proposed generalization of independent vector analysis IVA, achieves asymptotically the minimal mean square error MMSE in the separation of multidimensional Gaussian components.

Fast Independent Vector Extraction by Iterative SINR Maximization

  • Robin ScheiblerNobutaka Ono
  • Computer Science
    ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
  • 2020
This work proposes fast independent vector extraction (FIVE), a new algorithm that blindly extracts a single non-Gaussian source from a Gaussian background and finds that it is vastly superior to competing methods in terms of convergence speed, and has high potential for real-time applications.

Overdetermined blind separation for convolutive mixtures of speech based on multistage ICA using subarray processing

A new extended MSICA using subarray processing, where the number of microphones and that of sources are set to be the same in every subarray is proposed, which reveals that the separation performance of the proposed MSICA is improved as thenumber of microphones is increased.

Subspace-Based Channel Shortening for the Blind Separation of Convolutive Mixtures

A novel subspace-based channel shortening procedure is proposed based on the structure of the delayed autocorrelation matrices of the observation process which can be transformed into an instantaneous blind source separation (BSS) problem which is simpler to solve using, for example, independent component analysis techniques.
...