Neural Indicators of Deception: EEG Data Analysis for Lie Detection

This study investigates the neural markers of deception using EEG data and explores the efficacy of deep learning models in lie detection. A novel EEG dataset, "Lie Waves," was utilized, capturing brainwave activity during deceptive behavior. To classify deceptive states, both a Transformer model and a 1D Convolutional Neural Network were applied to the dataset. EEG signals were preprocessed and features were extracted using Discrete Wavelet Transform . The results demonstrated that the Transformer model outperformed the Convolutional Neural Network, achieving a classification accuracy of 92%, with superior precision, recall, and F1-score. This research highlights the potential of EEG-based deep learning models for efficient and reliable deception detection, offering significant improvements in accuracy over traditional methods. The findings contribute to advancements in biometric systems, particularly for lie detection applications.