Human Voice Localization in Noisy Environment by SRP-PHAT and MFCC

This paper proposes an efficient method based on the steered response power (SRP) technique for sound source localization using microphone array: the SRP-PHAT and MFCC. As compared to the previous SRP technique, the proposed method using MFCC is more efficient in localizing the human speech. The previous method focuses only on the loudest sound source in a given area and such sound source may also contain noises which have high spectral peaks than human speech. The proposed method will recognize the speech through vowel corpus comparison and enhance the speaker speech. And finally by using MFCC feature extraction technique the human speech can be extracted from the noisy environment and this speech signal is used for sound source localization using SRP-PHAT