Convolution Neural Network Efficiency Research in Gender and Age Classification From Speech
Anna V. Kuchebo, Vadim V. Bazanov, Igor A. Kondratev, Anastasia M. Kataeva · 2021
In this article, deep learning is investigated in the formulated problem of age and gender classification based on a person's speech. We conducted theoretical research of the effectiveness of the methods currently used to identify and verify a person by voice and explored methods of audio preprocessing. The system uses a software package that consists of several interconnected convolutional neural networks. Also, we used open data to train our classification algorithm. In conclusion, we conducted an experiment to analyze the effectiveness of the algorithm in the problem of age and gender classification.