Real-time Emotion Recognition System Based On EEG Signals

Emotion is one of the advanced functions of the human brain.Everyone can easily feel their emotions,but trying to detect the emotions of others is not an easy task.Accurate recognition of emotions though technical methods is one of the most active research hotpots in psychology and cognitive neuroscience.With the development of Brain Computer Interface(BCI)and the Artificial Intelligence(AI),the recognition of emotions based on EEG signals has become a new means of studying emotions.Firstly,EEG signals contain a lot of emotion-related information.Secondly,EEG signals have high temporal resolution and are not easy to camouflage.These two points make EEG show great advantages in real-time emotion recognition.In order to study real-time emotion recognition through EEG signals,this paper first designs and develops an EEG emotion experiment based on picture stimulation.Using images from IAPS as material,the emotional EEG signals of 22 subjects were induced and collected by this paper.After preprocessing the raw EEG signals,28 kinds of EEG features including time domain features,frequency domain features and nonlinear features were extracted by this research.After feature selection and classifier classification,the classification results were obtained in several different situations.Among them,the classification accuracy rate of 50% were obtained in the four emotions classification,the classification accuracy rates of 73.8% and 77.5% were obtained on the arousal and valance classifications,respectively.Through the discussion and analysis of the above classification results,the research explores the feasibility of emotion recognition based on EEG signals from the three leads of the forehead(Fp1,Fp2,Fpz).Based on the above conclusions,this paper develops a real-time emotion recognition prototype system based on forehead three-lead.Compared with the traditional emotion EEG recognition process,this prototype system enables real-time recognition of emotions quickly and easily,which proves it has certain application value.