Lightweight MATLAB GUI for Real-Time Voice Biometric Authentication

This paper presents a real-time, text-dependent voice biometric authentication system designed using MATLAB App Designer. The system utilizes Fast Fourier Transform (FFT) for feature extraction and cross-correlation for temporal matching, enabling efficient and accurate speaker verification. Tailored for environments with limited computing resources, the approach offers a streamlined alternative to complex techniques such as Mel-Frequency Cepstral Coefficients (MFCC). Operating in the frequency domain, the system achieves 95.3 % authentication accuracy. A user-friendly graphical interface supports seamless interaction, real-time user enrollment, and secure logging of verification attempts. With its focus on simplicity, speed, and effectiveness, the solution proves well-suited for practical deployment. Future improvements will explore the integration of deep learning models and adaptive noise suppression methods to enhance performance in dynamic, real-world conditions.