Maximum-likelihood acoustic source localization: Experimental results

In this paper, we show several experimental acoustic source localization results using the maximum-likelihood parametric method. A direct localization is obtained when the sound source is placed inside the convex hull of the array. In the far-field case, we demonstrate the localization via the cross bearing from several widely separated arrays. In the case of two sources, an alternating projection procedure is applied to the ML method to estimate the two DOA's from the observed data. The ML algorithm is shown to be effective in locating sound sources of various types, e.g., vehicle, music, and even white noise.