FIT2016 第15回情報科学技術フォーラム 開催日:2016年9月7日(水)~9日(金) 会場:富山大学キャンパス
抄録
J-034
A Measurement and Classification of Multiple Subjects' P300 for Image Retrieval
Junwei Fan・Hideaki Touyama(Toyama Prefectural Univ.)
P300 is a kind of event related potential. We can use P300 to detect the attention. This technique was applied to brain-machine interface for information retrieval, such as letters and images for single user. In this paper, we proposed an experiment which was the facial expression retrieval of the computer-generated images by using collaborative P300. As the results, a high classification performance was obtained. The mean F-measure value was 89.5%. We suggested that the proposed technique may be applicable to BMI for detecting events in future.