Online maximum-likelihood learning of time-varying dynamical models in block-frequency-domain

Malik, Sarmad Majeed; Enzner, Gerald

doi:10.1109/icassp.2010.5495841

Cited by 30 publications

(30 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The covariance parameters, i.e., and , of the CR-SSFDAF were learned by invoking the expectation-maximization-type learning rules described in [13] once per frame time . The value serves to maintain the tracking ability of the CR-SSFDAF.…”

Section: Simulation Resultsmentioning

confidence: 99%

“…Considering the DFT-domain CR state-space model given by (10) and (14), we can write the equations of the DFT-domain Kalman filter [12], [13] for the th channel as…”

Section: Cross-relation State-space Algorithmmentioning

confidence: 99%

“…In order to incorporate the time-varying nature of the unknown channels , we model as a random variable with a first-order Markov property [11], [13] …”

Section: B First-order Markov Modeling Of Time-varying Channelsmentioning

confidence: 99%

See 2 more Smart Citations

A State-Space Cross-Relation Approach to Adaptive Blind SIMO System Identification

Malik

Schmid

Enzner

2012

IEEE Signal Process. Lett.

Self Cite

View full text Add to dashboard Cite

In this work, we address blind single-input multipleoutput (SIMO) system identification in conjunction with dynamical modeling of the underlying system. A multichannel cross-relation observation model in the DFT domain is employed to derive a blind adaptive algorithm that recursively learns the posterior distribution on the unknown SIMO system. The proposed algorithm inherently incorporates the time-varying nature of the channels and a representation of the observation noise. We show that the resulting cross-relation state-space frequency-domain adaptive filter (CR-SSFDAF), owing to its stable and diagonalized structure and near-optimal step-size control, can be efficiently operated in time-varying and noisy conditions.

show abstract

Section: Simulation Resultsmentioning

confidence: 99%

“…Considering the DFT-domain CR state-space model given by (10) and (14), we can write the equations of the DFT-domain Kalman filter [12], [13] for the th channel as…”

Section: Cross-relation State-space Algorithmmentioning

confidence: 99%

See 1 more Smart Citation

A State-Space Cross-Relation Approach to Adaptive Blind SIMO System Identification

Malik

Schmid

Enzner

2012

IEEE Signal Process. Lett.

Self Cite

View full text Add to dashboard Cite

show abstract

“…As the proposed algorithm is based on the noisy linear frequencydomain echo path model described in [13] inference of its parameters introduced in [14], we give in the following section a short summary.…”

Section: Online Maximum-likelihood Learningmentioning

confidence: 99%

“…In [14] it is proposed to estimate both jointly with the adaptive filter coefficients by optimizing a single Maximum-Likelihood (ML) objective function. However, due to the high-dimensional linear Gaussian Discrete Fourier Transform (DFT)-domain state space model [13,14], this approach greatly overestimates the noise power after an abrupt system change occurred. This results in a slow reconvergence which has been analyzed theoretically in [15].…”

Section: Introductionmentioning

confidence: 99%

Noise-Robust Adaptation Control for Supervised Acoustic System Identification Exploiting a Noise Dictionary

Haubner

Brendel

Elminshawi

et al. 2021

ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

We present a noise-robust adaptation control strategy for blockonline supervised acoustic system identification by exploiting a noise dictionary. The proposed algorithm takes advantage of the pronounced spectral structure which characterizes many types of interfering noise signals. We model the noisy observations by a linear Gaussian Discrete Fourier Transform-domain state space model whose parameters are estimated by an online generalized Expectation-Maximization algorithm. Unlike all other state-of-theart approaches we suggest to model the covariance matrix of the observation probability density function by a dictionary model. We propose to learn the noise dictionary from training data, which can be gathered either offline or online whenever the system is not excited, while we infer the activations continuously. The proposed algorithm represents a novel machine-learning-based approach to noise-robust adaptation control which allows for faster convergence in applications characterized by high-level and non-stationary interfering noise signals and abrupt system changes.

show abstract

Online Acoustic System Identification Exploiting Kalman Filtering and an Adaptive Impulse Response Subspace Model

Haubner

Brendel

Kellermann

2022

J Sign Process Syst

View full text Add to dashboard Cite

We introduce a novel algorithm for online estimation of Acoustic Impulse Responses (AIRs) which allows for fast convergence by exploiting prior knowledge about the fundamental structure of AIRs. The proposed method assumes that the variability of AIRs of an acoustic scene is confined to a low-dimensional manifold which is embedded in a high-dimensional space of possible AIR estimates. We discuss various approaches which exploit a training data set of AIRs, e.g., high-accuracy AIR estimates from the acoustic scene, to learn a local affine subspace approximation of the AIR manifold. The model is motivated by the idea of describing the generally nonlinear AIR manifold locally by tangential hyperplanes and its validity is verified for simulated data. Subsequently, we describe how the manifold assumption can be used to enhance online AIR estimates by projecting them onto an adaptively estimated subspace. Motivated by the assumption of manifolds being locally Euclidean, the parameters determining the adaptive subspace are learned from the nearest neighbor AIR training samples to the current AIR estimate. To assess the proximity of training data AIRs to the current AIR estimate, we introduce a probabilistic extension of the Euclidean distance which improves the performance for applications with non-white excitation signals. Furthermore, we describe how model imperfections can be tackled by a soft projection of the AIR estimates. The proposed algorithm exhibits significantly faster convergence properties in comparison to a high-performance state-of-the-art algorithm. Furthermore, we show an improved steady-state performance for speech-excited system identification scenarios suffering from high-level interfering noise and nonunique solutions.

show abstract

Online maximum-likelihood learning of time-varying dynamical models in block-frequency-domain

Cited by 30 publications

References 6 publications

A State-Space Cross-Relation Approach to Adaptive Blind SIMO System Identification

A State-Space Cross-Relation Approach to Adaptive Blind SIMO System Identification

Noise-Robust Adaptation Control for Supervised Acoustic System Identification Exploiting a Noise Dictionary

Online Acoustic System Identification Exploiting Kalman Filtering and an Adaptive Impulse Response Subspace Model

Contact Info

Product

Resources

About