摘要
为研究视听双模态混合诱发脑机交互的有效性,设计单一听觉、单一视觉以及视听混合语义一致与否的诱发脑电采集实验,分别从行为学响应和事件相关电位ERP角度对不同诱发类型进行对比分析,提出一种基于空-频域的脑电信号特征提取方法,用于数字拼写分类识别。实验结果表明,视听双模态混合诱发的ERP幅值更大、范围更广,被试对语义一致的视听混合刺激反应更快,数字拼写分类准确率较高,将该方法与共空间模式特征进行比较,分类准确率提高了4.92%。
To study the effectiveness of bimodular audio-visual evoked brain-computer interaction,single auditory,single visual and audio-visual semantic coherence or not hybrid evoked EEG acquisition experiment was designed.Different evoked types were compared and analyzed from behavioral response and event-related potential(ERP)perspectives,and a space-frequency domain based EEG feature extraction method was proposed for classification and recognition of digital spelling.Experimental results show that the amplitude of ERP evoked by audio-visual bimodal is larger and range is wider.Moreover,the subjects respond more quickly to the semantic consistent audio-visual hybrid stimuli,and the accuracy of digital spelling classification is also hig-her.Compared with the common spatial pattern feature,the classification accuracy of this method is improved by 4.92%.
作者
郭柳君
张雪英
陈桂军
GUO Liu-jun;ZHANG Xue-ying;CHEN Gui-jun(College of Information and Computer,Taiyuan University of Technology,Taiyuan 030024,China)
出处
《计算机工程与设计》
北大核心
2020年第6期1755-1761,共7页
Computer Engineering and Design
基金
山西省应用基础研究基金项目(201701D221117)
太原理工大学校基金项目(2016QN24)。
关键词
视听混合
行为学
ERP分析
空-频域
特征提取
字符拼写
audiovisual hybrid
behavior
ERP analysis
space-frequency domain
feature extraction
character spelling
