Gaussian Mixture Models (GMMs) of power spectral densities of speech and noise are used with explicit Bayesian estimations in Wiener filtering of noisy speech. No assumption is made on the nature or stationarity of the noise. No voice activity detection (VAD) or any other means is employed to estimate the input SNR. The GMM mean vectors are used to form sets of over-determined system of equatio...

In this paper, we first analyze the problems of speech and noise contamination process in noise-masking point of view, and propose a new approach to estimate degree of noise masking effect on clean speech distribution model based on sequential noise estimation. Sequential noise estimation is performed frame-by-frame using interacting multiple model (IMM) algorithm, so that realtime implementati...

In this paper, we propose a noise robust speech recognition method by combination of temporal domain singular value decomposition(SVD) based speech enhancement and Gaussian mixture model(GMM) based speech estimation. The bottleneck of GMM based approach is a noise estimation problem. For this noise estimation problem, we incorporated the adaptive noise estimation in GMM based approach. Furtherm...

in this paper, for the purpose of investigating horizontal sound propagation, based on the results obtained from practical measurements in the strait of hormoz and available data on sound speed variations at different depths, a comprehensive model for shallow water multipath underwater acoustic channel is presented. mathematical modeling of multipath effect is based on ray theory and the image ...

Accurate estimation of Gaussian noise level is of fundamental interest in a wide variety of vision and image processing applications as it is critical to the processing techniques that follow. In this paper, a new effective noise level estimation method is proposed on the basis of the study of singular values of noise-corrupted images. Two novel aspects of this paper address the major challenge...

In this paper, a signal processing method is proposed to estimate the low and high-frequency impedances of power systems using several short-term low power signal injections for a frequency range of 0-150 kHz. This frequency range is very important, and thusso it is considered in the analysis of power quality issues of smart grids. The impedance estimation is used in many power system applicati...

The suggested open problem concerns the error of estimation and the minimum of the cost in the filtering and optimal control problems for a partially observable linear system corrupted by wide band noise processes. Recent results allow to construct a wide band noise process in a certain integral form on the basis of its autocovariance function and design the optimal filter and the optimal contr...

Kalman filtering is a well-established methodology used in various multi-sensor data fusion applications. In our experiment, we first obtain measurements from the accelerometer and gyroscope and fuse them using Kalman filter in an inertial measurement unit (IMU). We estimate Kalman filter output and estimation error. The affect of process noise and measurement noise on estimation error is teste...

In this report, we describe our study of effect of noise on parameter estimation for statistical machine translation. So far, no study has been done on this topic, even though the algorithm used for parameter estimation for statistical machine translation (the EM algorithm) is known to be highly sensitive to noise. We present in detail the experiments performed to observe the influence of noise...