ABSTRACT
Acoustic Source Localization (ASL) means to estimate the position of the source, emitting sound. It has a wide application area. Based upon the literature survey, ASL techniques have been broadly classified as Time Delay Estimation (TDE), Beamforming, High Spectral Resolution, Energy aware and Binaural based. There are different methods based upon these classifications such as Cross correlation, Generalized Cross Correlation (GCC) and Adaptive Eigen Value decomposition (AED). It has been shown by the researchers that different techniques have proved to be effective in different environment conditions. GCC and AED have better performance in ideal propagation scenarios or where the reverberation is low. In moderate to high reverberation channels, SRP-PHAT has been proved to give more accurate results than GCC but the Computational complexity has always been a challenge here. High spectral resolution methods are quite complex but have given better resolution over other techniques and mainly preferred for narrowband signals to find their Direction of Arrival (DOA).
