In this paper, a gain-adapted speech recognition in unknown noise is developed in time domain. The noise is assumed to be the colored noise. The nonstationary autoregressive (NAR) hidden markov model (HMM) used to model clean speeches, The nonstationary AR is modeled by polynomial functions with a linear combination of A4 known basis functions. Enhancement using multiple Kalman filters is perfo...

It is by now established that, remarkably, the addition of noise to a nonlinear system may sometimes facilitate, rather than hamper the detection of weak signals. This phenomenon, usually referred to as stochastic resonance, was originally associated with strictly periodic signals, but it was eventually shown to occur for stationary aperiodic signals as well. However, in several situations of p...

To overcome the limitation induced by the fixed timefrequency resolution over the whole time-frequency plane of Gabor frames, we propose a simple extension of the Gabor theory leading to the construction of frames with timefrequency resolution evolving over time or frequency. We describe the construction of such frames and give the explicit formulation of the canonical dual frame for some condi...

The mbuston scheme is a novel reduced-detail paradigm for nonstationary signal modelinglprocessing with enhanced statistical stability. Here, we apply the robuston scheme to the problem of detecting a nonstationary random signal in white Gaussian noise. We propose two different “robuston detectors” along with signal-adaptive online implementations that perform online estimation of the signal st...

This paper establishes several results for uniform convergence of nonparametric kernel density and regression estimates for the case where the time series regressors concerned are nonstationary null– recurrent Markov chains. Under suitable conditions, certain rates of convergence are also established for these estimates. Our results can be viewed as an extension of some well–known uniform consi...

In any problem involving images having scale-dependent structures, a key issue is the modeling of these multi-scale characteristics. Because multi-scale phenomena frequently possess nonstationary, piece-wise multi-model behaviour, the classic hidden Markov method can not perform well in modeling such complex images. In this paper we provide a new modeling approach to extend previous hierarchica...

In this paper, we consider the logistic regression model with an integrated regressor driven by a general linear process. In particular, we derive the limit distributions of the nonlinear least squares (NLS) estimators and their t-ratios of the parameters in the model. It is shown that the NLS estimators are generally not efficient. Moreover, the t-ratios for the level parameters have limit dis...

This paper considers the censored regression model under the assumption that the regressors are integrated. We show that Maximum Likelihood estimation is superconsistent and asymptotically mixed normal, implying that standard inference techniques remain valid, and that in general least squares estimation based on the positive observations only is superconsistent, but not mixed normal. An except...

∗Department of Mathematics, Technion-Israel Institute of Technology ∗∗Vologda State Pedagogical University and Vologda Scientific Coordinate Centre of CEMI RAS, Vologda, Russia email: [email protected] Received October 14, 2002 Evaluation of the rate of convergence of stochastic models, as time t → ∞, has been a subject of investigation by generations of probabilists. During the last two de...

This paper gives results on weak and strong convergence in L 2 (IR s) r of the cascade sequence (k;n) generated by the nonstation-ary matrix cascade algorithm k;n = jMj P j h k+1 (j) k+1;n?1 (M ?j); where for each k = 1; 2; : : : ; h k is a nite sequence of r r matrices and M is an integer dilation matrix. The limit as n ! 1 of the cascade sequence is an r-vector of functions (k) that is a solu...