top of page
  • Writer's pictureGSI

Mind-reading AI transforms brainwaves into text


Mind-reading AI transforms brainwaves into text
Mind-reading AI transforms brainwaves into text

From mind to text! The scientists at the University of Technology Sydney (UTS) have successfully utilized mind-reading AI, implementing the DeWave model, to convert brainwave recordings into written text. This practical application holds the promise of revolutionizing communication for individuals facing speech challenges and exploring applications in robotics, aligning with Sustainable Development Goal 9 (SDG 9) - Industry, Innovation, and Infrastructure.


Participants in the study wore sensor-equipped caps, capturing electrical brain activity. DeWave AI translated recorded electroencephalogram (EEG) data into text, exhibiting an initial accuracy of 40%, with ongoing research projecting an improvement exceeding 60%.

Charles Zhou, a crucial member of the UTS team, clarified that DeWave mastered interpreting brain signals by analyzing instances where signals correlated with specific sentences. The integration with a large language model (LLM), similar to ChatGPT, transformed DeWave into a 'brainy writer,' crafting sentences guided by identified signals.


Looking forward, researchers envision applications aiding those with speech impairments and contributing to robotics, addressing the objectives of SDG 3 (Good Health and Well-being) and SDG 9. The practicality of EEG, compared to MRI scans, enhances accessibility for potential widespread adoption.





This scientific accomplishment exemplifies how international collaboration among researchers can yield innovative solutions, reinforcing the belief in the strength of the international community. Science and technology, as demonstrated by this initiative, have the power to advance understanding, bridge divides, and serve the greater good.

 

 


More information: https://www.uts.edu.au

Youtube credits: @Thomas-HAI

bottom of page