SPEAKER IDENTIFICATION SYSTEM FOR GENDER RECOGNITION

Madelina, Kudam (2019) SPEAKER IDENTIFICATION SYSTEM FOR GENDER RECOGNITION. [Final Year Project Report] (Unpublished)

PDF (Please get the password by email to repository@unimas.my , or call ext: 082-583914/3973/3933) Madelina Kudam.pdf Restricted to Registered users only Download (2MB)

Abstract

Speaker Identification System is the process of identifying who is speaking from a set of known voices of speakers. It also one of the important systems in human-machine interaction. There are various applications used this system such as banking over telephone network, database access service, voice dialling, and voice mail. Gender Recognition through speech is the system to identify the gender of the speaker by analysing and processing a specific input speech signal. In this paper, Mel-Frequency Cepstral Coefficient (MFCC), Delta, and Delta-Delta are used for the pre-processing and feature extraction of the system. The speech data will be collected in the Malay language from 50 males and 50 females. After that, the classification methods; SVM, K-NN, and Naïve Bayes will be trained to determine the speaker gender and evaluate using a confusion matrix. In this study, Speaker Identification System is programmed using Python language. The purpose of this study is to investigate methods to train classifiers for Gender Recognition and evaluate the performance of the gender recognition on speaker identification system. From the results, the best classifier is the Naïve Bayes

Actions (For repository members only: login required)

View Item