Speaker identification using fuzzy logic

In these times one of the most important processes in speech processing is the classification of gender. Typically classification of gender is based on considering pitch as feature. The pitch value of male is lower than the female. In most of the topical research works the process of the classification of gender is performed using the aforementioned condition. In some cases the pitch value of female is lower and also pitch of some male is high, in that case the exact required result is not being produced by this classification method. By considering the aforesaid problem we have here suggested a new method for gender classification which considers three features. In this new method the gender of the speaker is identified by using fuzzy logic. Now by using the above three features training data set is generated to train fuzzy logic. The performance of the proposed technique in gender classification is shown by the implementation result.