Discriminating Fake and Real Smiles Using Electroencephalogram Signals With Convolutional Neural Networks

Genuineness of smiles is particular interest in the field human emotions and social interactions. In this work, we develop an experimental protocol to elicit genuine fake smile expressions on 28 healthy subjects. Then, assess type using electroencephalogram (EEG) signals with convolutional neural networks (CNNs). Five different architectures (CNN1, CNN2, CNN3, CNN4, CNN5) were examined differentiate between real smiles. We transform temporal EEG into normalized gray-scale images perform three-way classification classify smiles, neutral form subject-dependent classification. achieved highest accuracy 90.4% CNN1 for full spectrum. Likewise, accuracies 87.4%, 88.3%, 89.7%, 90.0% Beta, Alpha, Theta, Delta bands respectively. This paper suggests that CNNs models, widely used image problems, can provide alternative approach detection from physiological such as EEG.