Voice Gender Recognizer Recognition of Gender from Voice using Deep Neural Networks

Abstract

The proposed research work has used a Multilayer Perceptron (MLP) based deep learning model for deploying a gender-based user classification. The input to proposed model is a set of acoustics features, which was extracted by various packages. This work has developed a dataset consisting of 3168 data points, which were extracted from the recorded voice samples of both men and women. The voice samples are produced by using acoustics analysis. The proposed model is trained with different set of parameters and finally comes up with an MLP model that achieves an accuracy of 96 percent on the test dataset.

Topics & Concepts