Emotion Recognition Scheme via EEG Signal Analysis

Abstract

The recognition of brain states under different person and different task features has gradually become a research hotspot. While the challenges facing brain state recognition are low accuracy and time issues. In this paper, a emotion recognition scheme is proposed by analyzing the original EEG data. Firstly, we analyze the basic characteristics of the brain wave signal, and then use the attention based RNN algorithm to eliminate the various interference signals. Experiments show that compared with some mature brain wave recognition methods, the scheme improved the recognition accuracy and owns high efficiency.