From brain waves to words: Brain2Qwerty offers a new path to communication without surgery

Meta announced Brain2Qwerty v2, a non-invasive brain-to-text system using magnetoencephalography rather than surgical implants, developed with the Basque Center on Cognition, Brain, and Language and backed by a $5 million Digital Brain Project fund. It achieves 61% word accuracy overall (78% for the best participant), trained on roughly 22,000 sentences from nine volunteers who each wore the MEG device for 10 hours while typing, versus about 8% accuracy for prior non-invasive approaches. Meta released full training code for v1 and v2, and BCBL is releasing the v1 dataset.

Source: ai.meta.com ↗