Gender Recognition from Speech Signal Using 1-D CNN

Abstract

The Human speech contains paralinguistic information used in many speech-recognition applications like automatic speech recognition, speaker recognition and verification. Gender from voice is considered as one of the essential tasks to be detected for such applications. A set of relevant features are extracted from the speech signal to distinguish gender as female or male. This paper proposes a 1-D Convolutional Neural Network (CNN) model with different features like MFCC, Mel spectrogram and Chroma extracted from speech signal to recognize the gender. Experiments are carried on Mozilla voice dataset and evaluated performance of the model. The results show that the combination of MFCC, Mel and chroma feature sets give better accuracy of 97.8%.