Speech Enhancement Based On Masking Properties Of The Human Auditory System

Speech enhancement is becoming an important branch of speech signal process,which mainly used in the noise reduction, the preprocess of the speech recognition system and LPC. Single channel subtractive-type algorithms is often used in the speech enhancement, but musical residual noise will appear. In order to reduce the musical residual noise and the background noise, a speech enhancement method based on masking properties of the human auditory system is described. This method uses BARK wavelet packet transform to simulate the frequency feature of human auditory model to get the threshold. Experimental results have shown that this method provides speech of better clearness and intelligibility than conventional methods do. Two main achievements are got in this paper.1. Speech enhancement method based on masking properties of the human auditory system is described. In this paper, a threshold is got through simulating the frequency of human auditory in order to reduce the musical residual noise and the background noise.2.BARK wavelet is used in the speech enhancement.to decompose and reconstruct the speech. .Because BARK wavelet is coincide with human auditory system, the speech we got is more nature.