Throat Microphone for Speaker Recognition Using AANN

In this paper, we have analyzed the performance of speaker recognition system based on features extracted from the speech recorded using throat microphone in clean and noisy environment. In general, clean speech performs better for speaker recognition system. Speaker recognition in noisy environment, using transducer held at the throat results in a signal that is clean even in noisy. This speaker recognition system is also beneficial to visually impaired person. The system recognizes the speakers from acoustic features of mel-frequency cepstral coefficients (MFCC). AANN is one of the modeling techniques used to capture the feature. The auto associative neural network (AANN) is used to capture the distribution of the acoustic feature vectors in the feature space. This model captures the distribution of the acoustic feature of a class, and the backpropagation learning algorithm is used to adjust the weights of the network to minimize the mean square error for each feature vector. The experimental results show that, the performance of AANN using MFCC gives an accuracy of 94.93% in clean and noisy environment. Keywords— Autoassociative neural network(AANN); Mel-frequency cepstral coefficients(MFCC); Speaker Recognition(SR); Throat microphone; Visually impaired.