AI converts brain signals into text
Researchers from Meta, BSL University, and the Adolf De Rothschild Hospital have developed a non-invasive technique to convert brain activity into written text using a deep learning algorithm along with EEG and MEG technologies. This technology relies on measuring the electrical and magnetic activity of the brain and assists patients who are unable to speak or use electronic devices due to injuries or neurological disorders, all without the need for surgical sensor implants. The algorithm achieved an error rate of 29% with MEG technology, which was reduced to 18% with optimal performance, enabling high-accuracy decoding. This advancement offers hope for developing safe and effective assistive systems for people with disabilities.
Notice: This Is an AI-Generated Summary
Comments (0)