Facebook Building 8 officially unveiled the brain-computer interface research results

Regina Dugan, Deputy Director of Engineering for Facebook under Facebook, gave a speech today at F8. She emphasized some of Facebook’s research in the brain-computer interface field. Although it is still in its early stages, Facebook wants to solve the current AR input problem with a technology that "basically knows what you are thinking".

In six months, Facebook has formed a team of more than 60 people, including scientists specializing in machine learning for decoding speech and language, optical neuroimaging systems, and “the world’s most advanced neuroprosthetics”. Engineer, system integration developer. All of these talents are working hard to solve a problem: How do humans interact with the digital world without using sound and hands?

At first glance, this problem is well-suited for patients who lack limbs, such as patients with locked-in syndrome (a condition that causes general paralysis and speechlessness). But in consumer technology, what Dugan calls "the brain mouse of AR" (which lets you click binary "yes" or "no") may have a big impact on this area. Dugan said that his goal is to achieve direct brain-text input, and this is exactly the "smooth brain-computer interface required by AR."

Although research on brain-computer interfaces is mainly focused on the medical field, Dougan pointed out that the overall goal of the project is to develop a brain-computer system that allows you to input 100 words per minute (5 times faster than you type text on your smartphone). Reading the brain's language center directly. This can not only benefit people with disabilities, but it is also for everyone.

“What we are talking about is decoding these words. Before we say it, we have decided to send words to the brain language center: a silent voice interface with all the flexibility and voice speed, but also provides the privacy of typed text. Dugan said that this type of input can bring immeasurable value to always-on wearable devices such as AR heads-up.

Because the systems currently in use do not work in real time and require electrodes to be surgically implanted (a huge obstacle we have not yet crossed), the new Facebook team is investigating non-invasive sensors based on optical imaging. Dugan said that the device needs to sample data hundreds of times per second while being accurate to millimeters. The requirements are very demanding but technically feasible.

This can be achieved by bombarding the brain with quasi-ballistic photons, and Dugan says this lighter particle can read the brain more accurately. When designing a non-invasive system based on optical imaging, you need light to pass through the hair, skull and various tissues before reading the brain. Again, the project study is still in its early stages, but Facebook believes that optical imaging is the best entry point.

Dugan said that the overall goal is to break the barriers between languages ​​by reading the semantics behind humanistic languages ​​(such as Chinese or Spanish), thus creating ways to connect people.

MPPT Solar Charger Controller

Mppt Solar Charger Controller,Wind Solar Controller,Mppt Controller,Micro Solar Charge Controller

GuangZhou HanFong New Energy Technology Co. , Ltd. , https://www.gzinverter.com