Facebook is creating an interface tool for brain-computer Augmented Reality (AR) to assist users to type their minds.
The firm announced its Brain-Computer Interface (BCI) program at its F8 Developers’ Conference in 2017 — outlining its objective of building a non-invasive, wearable device that allows individuals to type by just imagining speaking.
Facebook supports a team of scientists at the University of California, San Francisco (UCSF) working to assist neurologically damaged patients talk again by identifying in a real-time intended speech from brain activity.
The UCSF team “shared how far we must go as a prospective input alternative to attain non-invasive BCI, in an article released in the journal Nature Communications,” says Facebook in a Tuesday blog post.
A small set of full, spoken words and phrases from brain activity were decoded in real-time by the UCSF team— a first in the field of BCI research.
Researchers emphasize that their algorithm can only recognize a tiny set of words and sentences so far, but continuing research aims at translating much bigger vocabulary with dramatically reduced error rates.
“AR’s promise lies in its capacity to link individuals seamlessly to the globe around them and to each other. Instead of looking down on a mobile screen or breaking out a laptop, we can keep eye contact and collect helpful data and context without ever missing a beat,” added Facebook.
“We stand on the brink of the new wave of human-oriented computing, where mixed technologies between RA and VR converge and revolutionize the interplay with the globe around us,” as Chief Scientist Michael Abrash and his crew at Facebook Realities Labs (FRL) view it.
“It’s going to be something completely new, as it’s going to wash a break from anything that’s happened before, as the interface based on the mouse / GUI was from punch cards, printouts, and teletype machines,” Abrash said.
The goal of Facebook Reality Labs ‘ BCI study program is to create a non-invasive, silent voice interface that will allow individuals to type just by imagining the phrases they want to say-a technology that could one day be a strong input for wearable all-day AR glasses.
The researchers ultimately hope to achieve a real-time decoding rate of 100 words per minute with a 1,000-word vocabulary and less than 17 percent word error rate.
Facebook first announced in 2017 that it was working on a computer-brain interface with its study lab, Building 8.
Neuralink’s courageous study, which has uncovered small brain “threads” in a chip that is long-lasting, usable at home and has the ability to replace cumbersome equipment presently used as brain-machine interfaces, comes on the heels of Elon Musk-led startup Neuralink.