Work

Fast Fourier Transform and Discrete Wavelet Transform to Enhance Biometric Voice Recognition

Year: 2018

Type: article

Abstract: This study investigates one of the technologies used in user recognition systems. Voice recognition is used as a tool to protect information systems from unlawful access. Other biometric techniques such as fingerprint, face print and iris print have shown some usage limitations such as big storage memory in the iris print technology. VRSs are based on the user's voice which is impossible to imitate and does not require costly facilities such as big storage memory. Using MFCC, DWT, and FTT, this study proposed an AVRS where the system trains and tests the voice of 10 speakers in two languages; English and Arabic. Previous studies have used only one language to train and test the speaker. The results of this study indicate that the rejection rate for Arabic words (12.9%) is higher than that for the English words (5.5%) which mean that the system is affected by the spoken words.

Source: مجلة البحوث الأكاديمية

DOI: 10.65540/jar.v12i.963

Authors: Najat Ahmed Elteraiki1, Mohammed M. Elsheh Mohammed M. Elsheh

Institutions: Libyan Academy, The Libyan Academy for Postgraduate Studies-Misurata

Language: English

Topic: Speech Recognition and Synthesis

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: hybrid

OpenAlex ID: W7128962016

FWCI:

Cites: 8

Cited by:

Related to: 10