计算机科学
脑电图
脑-机接口
频道(广播)
选择(遗传算法)
人工智能
噪音(视频)
运动表象
语音识别
变压器
模式识别(心理学)
机器学习
工程类
图像(数学)
电压
电信
心理学
精神科
电气工程
作者
Wei Mu,Junkongshuai Wang,Lu Wang,Pengchao Wang,Jiaguan Han,Lan Niu,Jianxiong Bin,Lusheng Liu,Jing Zhang,Jie Jia,Lihua Zhang,Xiaoyang Kang
标识
DOI:10.1109/bci57258.2023.10078658
摘要
EEG is widely used in the field of brain computer interfaces. It has the advantages of being non-invasive, easy acquisition, and so on. However, multi-channel EEG signals will bring noise interference and redundant information while improving the resolution, and channel selection can reduce noise signals to obtain more real signals. Therefore, appropriate channel selection is very necessary for BCI system application. This paper proposes a Fisher score calculation method based on OVR-CSP features for channel selection and uses a transformer to classify. The experimental results show that the proposed method achieves 85.54% accuracy in selecting 16.55 channels on the public dataset BCI IV IIa.
科研通智能强力驱动
Strongly Powered by AbleSci AI