Single channel speech enhancement technique — An implementation on MATLAB

Speech is one of the efficient communication medium; if it can be recognized by machine then human interaction can be improved. As the noise produced by various ambient sources like vehicles normally lies in this frequency range, speech signals get easily distorted by the ambient noises or AWGN. The degraded speech therefore needs to be processed for the enhancement of the speech components. The aim of the speech enhancement is to improve the quality and intelligibility of degraded speech signal. In this work, two types of noises are added to the speech signal, AWGN and Traffic noise. The noisy speech signal is applied with Spectral subtraction method and modified versions of spectral subtraction methods. The outcomes of technique are discussed in detail.