Over the previous years, a marvelous quantity of study was performed by utilizing the artificial intelligence based deep learning approaches for the gender recognition applications. The gender recognition facing the problems in as preprocessing, feature extraction and classification stages mostly through the speech input signals, thus solving these problems is mandatory to improve the classification accuracy of speech processing. To provide the prominent solution, this paper focuses on investigation of various speech recognition methodologies developed by the various researches in the past few years. Initially, spectrum subtraction method is used to perform the preprocessing of speech signal. Then, MFCC features are extracted from speech signal and tested with the Deep learning convolutional neural network (DLCNN) model for classifying the gender. The extensive simulation results shows that the proposed method gives the better classification accuracy compared to the state of art approaches.