Enhancing voice authentication systems with deep fake audio detection

Lawate, Atharva (2024) Enhancing voice authentication systems with deep fake audio detection. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (699kB) | Preview
Preview	PDF (Configuration Manual) Download (786kB) | Preview

Abstract

This Report initiates a broad study on the advance of voice authentication systems with state-of-the-art mechanisms for the detection of deepfake audio. Motivated by the tremendous threat from deepfake technology in the approach of voiceprint-based security systems, the research aims to estimate and advance the efficiency of the prevailing detection procedures. Using a dataset from Kaggle and carrying out the analysis of the LSTM network on the python platform, the research also makes use of feature extraction robustness and careful evaluation metrics that are followed rigorously. Some of the key findings include the high confidence and robustness of the LSTM model to identify whether the audio is real or synthetic. These make reasonable contributions to speech and audio security, intended to be quite promising for effective work and commercialization.

Item Type:	Thesis (Masters)
Supervisors:	Name Email McLaughlin, Eugene UNSPECIFIED
Uncontrolled Keywords:	LSTM; CNN; Deep learning; Voice authentication; machine learning
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > QA Mathematics > Computer software > Computer Security T Technology > T Technology (General) > Information Technology > Computer software > Computer Security Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Cyber Security
Depositing User:	Ciara O'Brien
Date Deposited:	30 Jul 2025 10:32
Last Modified:	30 Jul 2025 10:32
URI:	https://norma.ncirl.ie/id/eprint/8332

Actions (login required)

View Item