PERFORMANCE ANALYSIS OF SPEECH SIGNAL ENHANCEMENT TECHNIQUES FOR NOISY TAMIL SPEECH RECOGNITION

Abstract

Speech signal enhancement techniques have reached a considerable research attention because of its significant need in several signal processing applications. Various techniques have been developed for improving the speech signals in adverse conditions. In order to apply a good speech signal enhancement technique, an extensive comparison of the algorithms has always been necessary. Therefore, the performance evaluations of eight speech signal enhancement techniques are implemented and assessed based on various speech signal quality measures. In this paper, the Geometric Spectral Subtraction (GSS), Recursive Least Squares (RLS) Adaptive Filtering, Wavelet Filtering, Kalman Filtering, Ideal Binary Mask (IBM), Phase Spectrum Compensation (PSC), Minimum Mean Square Error estimator Magnitude Squared Spectrum incorporating SNR Uncertainty (MSS-MMSE-SPZC), and MMSE-MSS using SNR Uncertainty (MSS-MMSE-SPZC-SNRU) algorithms are implemented. These techniques are evaluated based on six objective speech quality measures and one subjective quality measure. Based on the experimental outcomes, the optimal speech signal enhancement technique which is suitable for all types of noisy conditions is exposed.

Keywords

• Geometric spectral subtraction (GSS)
• RLS adaptive filtering
• Wavelet filtering
• Kalman filtering
• Ideal binary mask (IBM)
• Phase spectrum compensation (PSC).