面向瘫痪失语者的脑机字符与语音交互系统  

Brain-Computer Character and Speech Interaction System for Paralyzed Aphasia

作　　者：余剑雄 文元美[1] Yu Jianxiong;Wen Yuanmei(School of Information Engineering,Guangdong University of Technology,Guangzhou 510006,China)

机构地区：[1]广东工业大学信息工程学院,广东广州510006

出　　处：《自动化与信息工程》2021年第3期1-8,39,共9页Automation & Information Engineering

摘　　要：为解决瘫痪失语者语言交流的困难,研究一种基于单通道柔性干电极脑电采集模块的脑机字符与语音交互系统。首先,将时序脑电信号绘制成二维图像并着色处理,利用卷积神经网络识别出含有眨眼特征的图像;然后,根据含有眨眼特征图像的连续帧数,实现短眨眼、中眨眼、长眨眼特征的细分;最后,结合脑电信号中解析的注意力特征,在虚拟交互系统输出字符与语音。实验结果表明:该系统的交互准确率达96.4%。To help paralyzed aphasia patients to solve the difficulty of language communication,a new brain-computer interaction system based on single channel flexible electrode EEG acquisition module was studied.Firstly,the sequential EEG signal is drawn into a two-dimensional image and colored.Secondly,the image with blink feature is recognized by CNN.Thirdly,according to the number of consecutive frames of the image with blink feature,the features of short blink,medium blink and long blink are subdivided.Finally,combined with the attention feature of EEG analysis,characters and speech are output in the designed virtual interactive system.The experimental results show that the interactive accuracy of this method is as high as 96.4%.

关 键 词：脑电信号采集 单通道 眨眼特征 注意力特征 字符与语音交互系统 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术]