Experimental research on emotion recognition based on brain-computer interface and brain waves

Abstract

Emotion recognition and classification are important research contents in the field of emotional computing. The current research focuses on the visual field and the speech field, but the accuracy of the emotion recognition and the classification which can be achieved so far is low, which is not enough for commercial applications. At present, due to the rapid progress in research on the brain waves and the brain-computer interfaces, and the great application value in the fields of the medicine and the military, this paper uses the brain electrode caps to collect the brain waves of the human brains under the seven different emotional states. The brain-computer interface transmits the brain wave patterns and the data to the computer, observes the brain waves in the OpenBCI_GUI graphical interface and records the changes in real time. After obtaining the brainwave data under the different emotional states, this paper uses the three statistical methods, such as the AdaBoosting algorithm, to perform the emotional classification on the recorded brainwave data. The experimental results show that the classification effect is good.