This paper proposes a new speaker localization method that is based on a preliminary estimation of the head orientation. The basic information on which the estimation is accomplished is called Oriented Global Coherence Field (OGCF).

The new algorithm is shown to be significantly more robust than the traditional ones so far explored. Its robustness is also due to an effective speech activity detection, implicitly performed by a thresholding technique applied to OGCF information. To show the performance of the proposed system, experiments were conducted on the NIST RT-05 Spring Evaluation source localization task, which is based on real recordings of lectures in noisy and reverberant environments.