Brain Interface Speaks Your Thoughts In Near Real-time

Published by admin on

Commentary, video, and a publication in this week’s Nature Neuroscience herald a significant advance in brain-computer interface (BCI) technology, enabling speech by decoding electrical activity in the brain’s sensorimotor cortex in real-time. Researchers from UC Berkeley and UCSF employed deep learning recurrent neural network transducer models to decode neural signals in 80-millisecond intervals, generating fluent, intelligible speech tailored to each participant’s pre-injury voice. Unlike earlier methods that synthesized speech only after a full sentence was completed, this system can detect and vocalize words within just three seconds. It is accomplished via a 253-electrode array chip implant on the brain. Code and the dataset to replicate the main findings of this study are available in the Chang Lab’s public GitHub repository.

Categories: Leben (Life aka misc)Technology

Related Posts

Leben (Life aka misc) Technology

Ring Says New Encryption Limits What It Can Give Police

Ring is rolling out a new default encryption system called TAKE, or “Throw Away the Key Encryption,” that rotates video keys every five minutes and permanently deletes Ring’s copy after 24 hours. The system is Read more…

Leben (Life aka misc)

OpenAI Is Developing a ‘Persistent’ AI Agent

Wired reports that OpenAI is testing a new “Persistent mode” for Codex that would let the agent keep working until explicitly “put to sleep,” proactively creating follow-up tasks for itself across sessions and using prior Read more…

Leben (Life aka misc)

Canada Hires 48 Scholars Away From Top US Universities

Canada is recruiting 64 researchers to universities across the country, including 48 from U.S. institutions such as Harvard, Yale, and MIT. The hires are backed by more than $364 million in government funding and are Read more…