A single micropohone kalman filter based noise canceller

A single micropohone kalman filter based noise canceller -...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
IEEE SIGNAL PROCESSING LETTERS, VOL. 6, NO. 3, MARCH 1999 55 A Single Microphone Kalman Filter-Based Noise Canceller M. Gabrea, E. Grivel, and M. Najim, Fellow, IEEE Abstract— A great deal of attention has been paid to speech enhancement using a single microphone system. The various approaches, based on the Kalman filter, operate in two steps: 1) the noise variances and the parameters of the speech model are estimated, and 2) the speech signal is retrieved using standard Kalman filtering. This letter presents an alternative solution that does not require the explicit estimation of the noise and the driving process variances. This deals with a new formulation of the approach proposed within a control literature framework by Mehra. Index Terms— Innovation, Kalman filter, speech enhancement. I. INTRODUCTION G IVEN a sequence of speech signal corrupted by an additive white noise, our purpose is to retrieve the speech signal. Several approaches have been reported in the literature. They essentially differ in the way they estimate both the speech model parameters and the noise variances. Kalman filtering for speech enhancement was proposed in [1], where speech parameters were obtained from the clean speech signal, before being contaminated by the noise. The method developed in [2] provides a suboptimal solution, which is a simplified version of the estimate-maximize (EM) algorithm based on the maximum likelihood argument. In [3] an adaptive algorithm which estimates the model parameters in a sequential way is proposed. The speech signal in [4] is modeled as an autoregressive moving average (ARMA) process, which is then estimated using a standard Kalman filter. In the methods mentioned above, the use of Kalman filtering always implies the estimation of the variances of both additive white noise and driving process. In this paper, we propose a Kalman filter based-method that avoids the explicit estimation of noise variances. For such a purpose, we reformulate the approach developed by Mehra in the field of identification [5], [6]. Unlike existing methods, no voice activity detector (VAD)
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 2
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 3

A single micropohone kalman filter based noise canceller -...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online