This paper presents an improved wavelet-based dereverberation method for automatic speech recognition (ASR). Dereverberation is based on filtering reverberant wavelet coefficients with the Wiener gains to suppress the effect of the late reflections. Optimization of the wavelet parameters using acoustic model enables the system to estimate the clean speech and late reflections effectively. This ...

The paper discusses two algorithms for solving the Zakai equation in the time-homogeneous diffusion filtering model with possible correlation between the state process and the observation noise. Both algorithms rely on the CameronMartin version of the Wiener chaos expansion, so that the approximate filter is a finite linear combination of the chaos elements generated by the observation process....

Motivated by the recent work of Benjamini, Häggström, Peres, and Steif (2003) on dynamical random walks, we: (i) Prove that, after a suitable normalization, the dynamical Gaussian walk converges weakly to the Ornstein–Uhlenbeck process in classical Wiener space; (ii) derive sharp tailasymptotics for the probabilities of large deviations of the said dynamical walk; and (iii) characterize (by way...

DE algorithm is a population-based heuristic global search technology. The algorithm has simple principle, fewer control parameters, but has strong robustness, and good optimization performance. This paper uses differential evolution algorithm for parameters identification of Wiener model. Firstly, we analyze the influence of mutation rate F on global parallel search ability and convergence in ...

In this paper we study the Wiener–Hopf factorization for a class of Lévy processes with double-sided jumps, characterized by the fact that the density of the Lévymeasure is given by an infinite series of exponential functions with positive coefficients. We express the Wiener–Hopf factors as infinite products over roots of a certain transcendental equation, and provide a series representation fo...