Automated Multimodal Lie Detection using Deep Neural Networks | A late-fusion deep learning framework combining MFCC audio features and ResNet-18 visual micro-expression analysis on the DOLOS dataset (82% Accuracy).
pytorch feature-extraction convolutional-neural-networks mfcc affective-computing facial-expression-recognition speech-analysis multimodal-deep-learning resnet18 deception-detection late-fusion lie-detection automated-lie-detection deceptive-behavior micro-expression-analysis
-
Updated
Aug 6, 2026