Table 3 Deception Detection using Bag of Lies Dataset with Video, Audio, EEG, and GazeModalities
Bag of lies dataset | Features extracted | Classification |
|---|---|---|
Video | LBP features from 20 frames per video | SVM, Random Forest Multilayer perceptron |
Audio | Zero crossing rate, Spectral centroid, Spectral bandwidth, Spectra l roll-off, Chroma frequencies, and MFCC | Random Forest KNN Classifier |
EEG | 100 points from each channel | Random Forest, Multilayer perceptron, EEGNet |
Gaze | Fixation, Blinks, Pupil Size | Random Forest Multilayer perceptron |