Performance Comparison of Adaptive Filters with Kalman Filter for Speech Enhancement

The goal of this paper is to perform a performance analysis for speech enhancement using classical adaptive filters with kalman filter. In many applications like speech recognition, hearing aids, forensic applications and telephone conversations etc... As the enhancement of speech signals is of very important. The performance of the speech recognition system is also reduces if the speech signal is corrupted by noise. To remove the noise present in the speech signal, the adaptive filters shown the good improvement in increasing the Signal to Noise Ratio (SNR) values. The simulations are done using NOIZUES speech corpus for different SNR values using Least Mean Square (LMS), Normalized Least Mean Square (NLMS), Recursive Least Squares (RLS) and Kalman filter. From the results it is observed that kalman filter has shown good improvement in speech enhancement when compared to the other methods.