Application of convolutional neural network for gender and age group recognition from speech

One of the major long-standing challenges in the task of sound recognition is the ability to recognize gender and age group from human speech. Although research projects in this subject achieve noticeable results but the accuracy is not as high as expectation. Therefore this paper aims at building one convolutional neural network-based model to classify gender and age group from human speech. We evaluate accuracy of prediction of our model on three audio sets and achieve impressive results compared with some other ones. In addition, we compare results of our study with several published articles.