An advanced EEG signal based emotion detection using machin learning and deep learning technique

Using EEG data, for the first time, researchers were able to distinguish between real emotional responses and those that were fabricated. Real and synthetic facial emotions have been recorded for the first time using EEG recordings of people’s brainwaves (available here). Real smiles include genuine grins, phoney or produced smiles, and neutral expressions. There are three basic qualities of three EEG emotional expressions that may be obtained: a real, unbiased, and false or played grin. We used DWT and Empirical Mode Decomposition (EMD) DWT-EMD& it is across 3 frequency bands to extort evaluation of EEG from the Data. For testing the proposed strategies, several classifiers, such as KNNs, ANNs, & SVMs, were utilized. Each of our 28 participants was allocated one of three groups: genuine, unbiased, or fraudulent. It was shown that employing both DWT and EMD together produced better outcomes than using either DWT or EMD separately. There were distinctive differences in the spectral power characteristics of DWT, EMD, and DWT-EMD in each of the three emotional expressions, indicating that the brain was using diverse patterns across all frequency bands. Both DWT and EMD can properly classify emotions, classifiers, and crew members in binary classification trials. When used in combination with an ANN, DWT-EMD obtained 94.3 and 84.1 percent classification accuracy in the alpha/beta bands respectively. DWT-EMD should be used in conjunction with alpha and beta frequency areas in future studies because of our findings.