Estimation of Room Acoustic Transfer Function using Speech Model

This paper presents a sound source localization method using only a single microphone, where the GMM (Gaussian Mixture Model) of clean speech is introduced to estimate the acoustic transfer function from a user's position. The new method is able to estimate it without measuring impulse responses. The sequence of the acoustic transfer function is estimated by maximizing the likelihood of training data uttered from a position, where the cepstral parameters are used due to effectively represent useful clean speech. Using the estimated sequence data, the GMM of the acoustic transfer function is created to deal with the influence of a room impulse response. Then, for each test data, we find a GMM having the maximum-likelihood from among the estimated GMMs corresponding to each position. Its effectiveness is confirmed by talker direction experiments in a room environment.