Low SNR robust Chinese tone extraction based human auditory model

This paper proposes a robust Chinese tone extraction algorithm based on the human auditory mechanism and short-term stationarity of Chinese speech. In this method, we use the pooled-correlogram based on human auditory model to extract the pitch of speech. An unsupervised lateral inhibitory network is used to get the peak position, which simulates the lateral inhibitory phenomenon in the human auditory system. The pitch restriction between successive frames of speech is imposed to get rid of a miscarriage of justice in the output of the lateral inhibitory network. As shown in the experiments, the method can extract Chinese tone quite well even in rather low SNR cases. It can separate the individual tone clearly as two speakers talk simultaneously.