An adaptive perception based time-frequency analysis model

A model for time-frequency analysis based on adaptive perception function is given to simulate the auditory peripheral. This model is a direct extension of the ear wig distribution (EWD) that proposed by O'Donovan. The main improvement of the new model lies in the fact that the cascade compressive gammachirp (Cascade cGC) model for the gammatone model is substituted to construct the spectral window, and the ability of the signal's power adaptation is implemented. So, the distribution of the new model description is more accordant with auditory perception than that of the EWD model. In the end, computational examples applied to several typical signals are given to demonstrate the effectiveness for the proposed approach.