Meta's Brain2Qwerty Achieves 61% Accuracy in Decoding Typed Text
Meta has unveiled a non-invasive system that reconstructs typed text from brain activity with 61% word-level accuracy, significantly outperforming prior methods.
Meta has introduced a new system called Brain2Qwerty v2 that can reconstruct typed text from brain activity without surgery. The company reports that the system achieves a word-level accuracy of 61%, compared to roughly 8% for previous non-invasive methods.
The model was trained on approximately 22,000 sentences. Nine healthy volunteers participated in the experiment, each spending 10 hours in a magnetoencephalography scanner while typing heard phrases on a keyboard.
This breakthrough could have implications for assistive technology and human-computer interaction. The technology is still in early research stages.
Source: ForkLog