Intelligent emotion recognition system using electroencephalography and active shape models

Human emotion recognition has become one of the key steps towards advanced human-machine interactions. Brain waves or Electroencephalography (EEG) is one of the frequently used bio signals in emotion detection as it is found that the signal measured from the central nervous system has a relationship between physiological changes and emotions. Using facial expressions is another mode that could be used for emotion recognition using external physiological signals. This project investigates the possibility of identifying emotions using brain signals and facial expressions. EEG feature extraction is done, using Relative Wavelet Energy calculation and Discrete Wavelet Transform methods for feature extraction, and Artificial Neural Network for emotion classification. For facial feature extraction Active Shape Model is used while the facial emotion classification is done using a Support Vector Machine. The solution could be used to study about the behaviour of EEG signals as well as facial expressions in different mental states.