Using only a sensor-filled helmet combined with artificial intelligence, a team of scientists has developed a system that can translate brainwaves into written text. The technology, known as DeWave, uses an electroencephalogram (EEG) to record brain electrical activity through the scalp while participants read passages of text. These recordings are then converted into text using the AI model. Although the system’s accuracy is currently around 40 percent, recent data shows an improved accuracy exceeding 60 percent. The researchers believe that with further refinement, this technology could revolutionize communication for people who have lost speech, such as stroke patients, and have applications in robotics.
The system’s non-invasive and portable nature makes it an accessible tool for converting thoughts into written words. Unlike previous methods that relied on MRI scans, the use of EEG makes the process much more practical as subjects do not have to remain still inside a scanner. The DeWave system was trained by analyzing examples where brain signals matched specific sentences, enabling it to understand how these signals relate to words and sentences. By connecting DeWave to a large language model (LLM), the team was able to create sentences based on the EEG data. Both DeWave and the language model were then trained together to improve sentence writing based on the brain signals.
Craig Jin from the University of Sydney commended the work done by the team, calling it excellent progress. He noted the significant improvement in accuracy compared to previous attempts at converting EEG to text. This technology holds great potential for individuals who have lost the ability to speak, providing a means of communication. Furthermore, it could find applications in the field of robotics.
While there is still work to be done to optimize the system, the researchers are hopeful about its future implications. As the accuracy continues to improve, the potential benefits for individuals with speech impairments become increasingly significant. By harnessing the power of AI and neural interfaces, this technology may open up new possibilities for communication and enhance the quality of life for many people.
