Feature Extraction Using Empirical Mode Decomposition of Speech Signal

摘要

Speech signal carries information related to not only the message to be conveyed, but also about speaker, language, emotional status of speaker, environment and so on. Speech is produced by exciting the time varying vocal tract system with a time varying e xcitation. Each sound is produced by a specific combination of excitation and vocal tract dynamics. This paper presents a speaker identification system using empirical mode decomposition (EMD) feature extraction method. The EMD is an adaptive multiresolution decomposition technique that appears to be suitable for non linear, non stationary data analysis. The EMD sifts the complex signal of time series without losing its original properties and then obtains some useful intrinsic mode function (I MF) components. T he FFT is the most useful method for frequency domain feature extraction . Wavelet transform(WT) is yet another method for feature extraction.