Language independent voice-based gender identification system
Jisha C. Thankappan, Sumam Mary Idicula · 2010
This paper presents a solution to the sub-problem of the utterance classification problem in which speech samples are classified on the basis of the gender of the speaker. This can help a lot in the current speech recognition applications and telecommunication systems, where identifying the gender of the speaker before recognition process starts, can enhance the overall system performance to a great extent. This paper presents a gender identification algorithm which uses both fundamental frequency and MFCC as features and a Support Vector Machine with Multi Layer Perceptron kernel function for classification.