Formant Frequency Estimation using Computation auditory Scene method for Speaker Identification and Recognition

Abstract

Speaker recognition is the process which is used to identify a particular person. A human voice is one of the medium using which any person can be recognized. Human voice has many characteristics which may vary for individuals. Speaker identification or recognition can been done by analyzing individual voice signal hence it is commonly referred as  speech recognition. However; voice signal is one of the most common factors which can be copied easily. Copying the voice signal of one person is known as mimic and many mimicry artists can mimic many famous personalities. Mimicry is done for the entertainment purpose and it is popular worldwide. However its dark side may raise various unethical activities. Hence speech recognition is a sensitive matter which gives a broad area to the researchers so that authenticity may be maintained. Many research have been done in this field however existing techniques does not have 100% efficiency and many more research can be done to achieve better results. In this paper  formant frequency estimation is done using computation auditory scene technique for the purpose of speaker identification and recognition. The result shows that proposed technique performs better in comparison with the other existing techniques.