Frequency domain binaural model with front-back discrimination capability using artificial neural network

Abstract

Conventional binaural model or binaural hearing assistance systems have a well known ambiguity in front-back discrimination which is called as "front-back confusion" or "cone of confusion" in psychoacoustics. It is known that spectral cue of sound provides keys to solve this confusion in binaural listening condition and the peaks and notches of spectral components play main role to estimate the vertical angle in sagittal coordinate. In this paper, a frequency domain binaural model with front-back discrimination method using artificial neural model is proposed and examined for various HRTF catalogs. The performance of the model is examined for simulated conditions using various HRTF catalogs and the results show similar discrimination capability even if the learning process should be done for each catalogs. The experimental results using a dummy head is also provided.