시각정보와 인지정보간의 연관관계 분석 모델 설계

Abstract

Recently, with the development of artificial intelligence, DNN (Deep Neural Networks) has been developed, which greatly improves the visual classification performance of the machine. However, it is not still sufficient to keep up with the human’s vision level. In this paper, after defining the correlative architecture between DNN for vision information learning and DNN for BCI information learning, we designed an encoder-decoder algorithm that transforms visually-responsive EEG signals into meaningful images. This model is used to analyze EEG signals based on DNN which generates images depicting emotions similar to those seen by humans, and analyze the co-relationship between visual and brain cognitive information.