This study sought to recognize the identity of individuals from their sounds using artificial neural network (ANN) and support vector machine (SVM) algorithms. The speaker recognition was text-independent, which is an exceedingly complex task. Typically, attempts at speech recognition are narrower than other areas of speech processing. As the research innovation, speech recognition was accomplished using multiple features extracted from the acoustic signals. The classification was based on the ANN algorithm and the SVM classifier. Notably, the whole process was evaluated using grid data. In this study, a new text-independent speech recognizing system was proposed. The features selected included Mel-Frequency Cepstral Coefficients (MFCC), energy, and first- and second-order derivative features. The proposed system worked based on the Perceptron neural network and the SVM. At the best state, the system's recognition accuracy for 15 speakers (10 males and five females) was 99.2%.
