P
Petr Tichavsky
Researcher at Academy of Sciences of the Czech Republic
Publications - 100
Citations - 4219
Petr Tichavsky is an academic researcher from Academy of Sciences of the Czech Republic. The author has contributed to research in topics: Blind signal separation & Tensor. The author has an hindex of 27, co-authored 97 publications receiving 3839 citations.
Papers
More filters
Journal ArticleDOI
Posterior Cramer-Rao bounds for discrete-time nonlinear filtering
TL;DR: A mean-square error lower bound for the discrete-time nonlinear filtering problem is derived based on the van Trees (1968) (posterior) version of the Cramer-Rao inequality and is applicable to multidimensional nonlinear, possibly non-Gaussian, dynamical systems.
Journal ArticleDOI
Efficient Variant of Algorithm FastICA for Independent Component Analysis Attaining the CramÉr-Rao Lower Bound
TL;DR: An improved version of the FastICA algorithm is proposed which is asymptotically efficient, i.e., its accuracy given by the residual error variance attains the Cramer-Rao lower bound (CRB).
Proceedings ArticleDOI
Efficient variant of algorithm fastica for independent component analysis attaining the cramer-RAO lower bound
Zbynek Koldovsky,Petr Tichavsky +1 more
TL;DR: An improved version of algorithm FastICA is proposed which is asymptotically efficient, i.e., its accuracy attains the Cramer-Rao lower bound provided that the probability distribution of the signal components belongs to the class of generalized Gaussian distribution.
Journal ArticleDOI
Fast Approximate Joint Diagonalization Incorporating Weight Matrices
Petr Tichavsky,Arie Yeredor +1 more
TL;DR: A new low-complexity approximate joint diagonalization (AJD) algorithm, which incorporates nontrivial block-diagonal weight matrices into a weighted least-squares (WLS) AJD criterion, is proposed, giving rise to fast implementation of asymptotically optimal BSS algorithms in various scenarios.
Journal ArticleDOI
Near-field/far-field azimuth and elevation angle estimation using a single vector hydrophone
TL;DR: A new underwater acoustic eigenstructure ESPRIT-based algorithm that yields closed-form direction-of-arrival (DOA) estimates using a single vector hydrophone that significantly outperforms an array of spatially displaced pressure hydrophones of comparable array-manifold size and computational load but may involve more complex hardware.