Improvement of teleconferencing systems with microphone arrays

The most important problem for microphone systems? applications like speech recognition, audio recording, voice recognition is distortion of sound signal caused by environmental conditions. Nowadays, to obtain more powerful sound signals, microphone arrays are used by combining more than one microphones. There are mainly two advantages of microphone arrays over normal microphones; the necessary of being close the microphone has cancelled and the estimation of sound source location or talker tracking by using the difference between time delay arrivals of each microphone.For the microphone arrays, the generalized cross correlation (GCC) technique is used in that the time delays of arrivals are analyzed by accepting that the locations of microphones has been known. While using this technique, to obtain better quality of sound signals, beamforming process should be performed. In this study, Steered Response Power (SRP) has been used as beamforming technique by combining GCC technique with Phase Transform (PHAT) weighting function.In this study, the locations of simultaneous (with 75%) five talkers are tried to be find in 3D space. It is aimed to find the sound source locations of source signals using by microphone arrays. SRP-PHAT method is used to beamforming application through the source locations directions with the help of microphone arrays. The most powerful output signals are accepted as the location of talkers. In investigation, genetic algorithm is used as optimization technique.