Deception Analysis Using Deep Learning Based on Voice Stress Detection

Issac, Geethu (2022) Deception Analysis Using Deep Learning Based on Voice Stress Detection. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration manual) Download (1MB) | Preview

Abstract

Voice Stress Detection concurrently is a sorcery that targets to deduce deception calculated by identifying the amount of stress in the voice signal. It becomes conceivable to detect the stressed voice in this century with the significant development, Artificial Intelligence (AI). Voice, being the core for communication is a good source of input signal to an AI model to analyze deception. The demand for healthy mental life of this era is the prime objective tried to be fulfilled with this work. The difference in the fluency of speech of a stressed person from that of an unstressed using the Deep Learning method of Convolutional Neural Network (CNN) is the featured sweep of this work. The dataset used for the implementation of the CNN model for analysing deception is The Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS). RAVDESS is a combination of unisexual voices of 24 different subjects with six emotions, which was analysed using the neural network and later binary classified into stressed or unstressed. The CNN model is implemented at the beginning on the voice of a single actor followed by 24 actors. A comparison on the existing Machine Learning models with Deep Learning model was also performed. An accuracy of 72.5% was obtained in classifying voice with an acceptable percentage of true positives with the CNN.

Item Type:	Thesis (Masters)
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning B Philosophy. Psychology. Religion > Psychology > Stress (Psychology)
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Tamara Malone
Date Deposited:	26 Jan 2023 16:46
Last Modified:	03 Mar 2023 11:17
URI:	https://norma.ncirl.ie/id/eprint/6139

Actions (login required)

View Item