Brain–computer interfaces (BCIs) measure brain waves to assist communication and operation of equipment. However, for application of BCIs in real life settings, portability and environmental adaptability are crucial. For BCIs using visual stimuli, portability can be achieved by using augmented reality (AR) with head-mounted displays, but prior placement of markers in the physical space is required for displaying appropriate choices. In this study, we demonstrate an environmentally adaptable AR-BCI that uses machine learning and depth sensors. The virtual marker that the user is focusing on through the transmissive AR display is estimated from the user's electroencephalograms (EEG), and preliminary accuracy exceeds chance level. These results suggest that the use of machine learning and AR head-mounted displays can increase the adaptability of BCIs. © 2022 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.

中文翻译：

---

具有空间意识的增强现实脑机接口

脑机接口 (BCI) 测量脑电波以协助设备的通信和操作。然而，对于 BCI 在现实生活中的应用，便携性和环境适应性至关重要。对于使用视觉刺激的 BCI，可通过使用带有头戴式显示器的增强现实 (AR) 来实现便携性，但需要事先在物理空间中放置标记以显示适当的选择。在这项研究中，我们展示了一种使用机器学习和深度传感器的环境适应性 AR-BCI。用户通过透射式 AR 显示器关注的虚拟标记是根据用户的脑电图 (EEG) 估计的，初步准确度超过了概率水平。这些结果表明，使用机器学习和 AR 头戴式显示器可以提高 BCI 的适应性。© 2022 日本电气工程师学会。由 Wiley Periodicals LLC 出版。