New AI can give silent people voice
New AI can give silent people voice
California researchers created an AI-powered device that allows paralyzed patients converse in real time. A major advance in brain-computer interface (BCI) technology.
Works How
The system was developed by Berkeley and San Francisco scientists. Brain-sensing devices assess motor cortex activity, which governs speech. AI converts these impulses into speakable words.
The technology uses brain sensors on the brain and face muscles. The device can produce speech in one second when a person speaks, quicker than previous methods.
AI training
Researchers had patients pronounce sentences while the device collected their brain signals to teach the AI. They utilized pre-paralysis recordings of patients’ voices to make speaking seem more genuine.
The speech is more flowing and comprehensible than with prior BCI systems. It’s a major step toward natural communication, even though it’s not perfect.
Why It Matters
This might change the lives of ALS and severe paralysis patients. It lets individuals communicate better with caretakers, friends, and others than with previous technology.
A UC Berkeley professor engaged in the Nature Neuroscience paper, Gopala Anumanchipalli, said the system is fast and intuitive like Alexa or Siri but geared for non-speakers.
Faster and more expressive words are the team’s goals for the system. To enable speech-impaired persons connect with others, the tool will be improved.
