摘要
文章将增强现实技术、基于人工智能的目标识别技术与BCI相结合,通过采用微软的Hololens作为视觉刺激呈现平台,实时显示前端视频采集设备获取的场景信息,利用YO-LO检测场景目标,根据目标数量映射相应数量闪烁刺激色块,受试者利用SSVEP对检测到的目标进行敏感目标筛选。8名受试者进行相应试验,并成功诱发出明显SSVEP信号,利用1.5s数据提取4种频率特征,分别获得90.00%,90.00%94.00%和90.00%的平均识别率,在敏感目标在线检测试验中获得很好的效果。该研究表明,本系统为敏感目标筛选提供了新的技术方案,并为BCI系统的日常化应用提供了可能性。
This research combines augmented reality technology,artificial intelligence-based target recognition technology and BCI.By using Microsoft's Hololens as a visual stimulus presentation platform,real-time display of scene information obtained by front-end video acquisition equipment,and detection of scene targets using YO-LO,the corresponding scintillation color patches were mapped according to the number of targets,and subjects used SSVEP to screen sensitive targets for detection.Eight subjects carried out corresponding tests,and suc-cessfully induced significant SSVEP signals.Four frequency characteristics were extracted by 1.5 s data,and the average recognition rate of 90.00%,90.00%,94.00%and 90.00%were obtained,respectively.Good results were obtained in the online detection test of sensitive targets.This study shows that this system provides a new technical solution for sensitive target selection,and provides a possibility for daily application of BCI sys-tem.
作者
马留洋
蒋龙杰
王宁
MA Liu-yang;JIANG Long-je;WANG Ning(The 27th Research Institute of China Electronics Technology Group Corporation,Zhengzhou 450047,China)
出处
《电光系统》
2024年第2期33-39,共7页
Electronic and Electro-optical Systems
关键词
脑-机接口
稳态视觉诱发电位
增强现实
人工智能
YOLO
Brain-Computer Interface
Steady State Visual Evoked Potential
Augment Reality
Artificial Intel-ligence
YOLO
