Klasifikasi Gender Melalui Suara Menggunakan Jaringan Propagasi Balik

Abstract

Voice recognition is a tery complicaled thing for a computel , and so we need an application that can distinguish the voices of male and female. In this srudy the authors use a baclc propagation neural networb to classify genfur with voice. Sampled onalog toice sigtals into digital signals .with the help of sofiware Adobe Audition , and then processed by the method of Linem predicive Coding ( LpC ) to obtain LPC cofficients . The LPC coeffcients are processed by iast Fowier Transform ( FFT ) and Chirp Z Tmnsform ( CZT ) to obtain new and better signal. The sign/ output of CZT is /urther processed by back propagation ANN to classifl gender through sound. Training dato used 25 male vo{ce and 25 female voice , and test dflta consisted of 5 male voice and 5 female voice , test results obtained from this study is 90 %.