SOURCE RECORD / S0200
基于卷积神经网络与度量学习的跨受试者可穿戴 P300 脑机接口
Subject-Independent Wearable P300 Brain-Computer Interface Based on Convolutional Neural Network and Metric Learning
- 作者或机构
- Hu L, Gao W, Lu Z, Shan C, Ma H, Zhang W, Li Y
- 发布日期
- 2024-09-25
- 来源类型
- peer_reviewed_human_dataset_study
- 技术路线
- non-invasive wearable EEG P300
- 功能场景
- assistive spelling and low-calibration BCI
- 权利状态
- manual_review / CC BY-NC-ND 4.0 (association not snapshotted)
证据摘要
证据支持：跨受试者可穿戴 P300 解码和少校准方向。关键数字：无校准为 73.23+/-7.62%，微调后为 78.75+/-6.37%。局限：准确率不是字符每分钟或每小时误选；没有报告患者、跨日、居家、商品化和中国总拥有成本证据。
关键数字
73.23+/-7.62% without calibration; 78.75+/-6.37% with fine-tuning
局限
Accuracy is not characters per minute or errors per hour; patient, cross-day, home-use, product availability and China TCO evidence not reported