Efficient modeling of the head-related transfer functions

Abstract

Efficient representations of the head-related transfer functions (HRTFs) based on autoregressive moving average (ARMA) modeling are explored in this chapter. A relatively new technique for ARMA order estimation, based on the minimum eigenvalues of a covariance matrix, is used to estimate efficient model structures. A major advantage of this method is that no prior estimation of the model parameters is required. Examples will be given which indicate that the HRTFs are primarily autoregressive (AR) systems.