AI-Powered Interface Allows Users to Control Objects with Their Minds

Researchers from Carnegie Mellon University's He Lab have made significant progress in advancing noninvasive brain-controlled interfaces. Using AI technology, the researchers have enhanced the decoding of human intention, allowing users to control a moving virtual object just by thinking about it. This achievement demonstrates unmatched performance in noninvasive brain-computer interfaces (BCIs). Noninvasive BCIs have advantages over invasive BCIs like Neuralink or Synchron, including increased safety and cost-effectiveness. However, they come with challenges such as less accurate recordings. The researchers conducted a study involving 28 participants and used AI and a deep neural network to decode their intentions from brain activity data recorded through electroencephalography (EEG). This AI-powered noninvasive BCI has the potential to assist motor-impaired patients and is being tested in able-bodied individuals and stroke patients. --- Send in a voice message: https://podcasters.spotify.com/pod/show/tonyphoang/message