Emotion Recognition Based On CNN

Emotion is a state that comprehensively represents human feeling, thought, behavior and it exists everywhere in daily life. Emotion recognition is an important interdisciplinary research topic in the fields of neuroscience, psychology, cognitive science, computer science and artificial intelligence. Neural network is a statistical learning model inspired by biological neural networks. This paper attempts to use the EEG signal from the DEAP data set to classify the emotion of the subjects, this data set represents the emotional classification research. Then the principal component analysis is used to reduce the dimension of the preprocessed EEG data, so the main emotional EEG features are obtained. Then the accuracy of the classification of the training samples and the test samples is tested by the CNN algorithm, and the other classification methods are compared to obtain the nerves. The network can be used as a robust classifier for brain signals even better than traditional learning techniques.