BOFFINS HAVE DEVELOPED an artificial intelligence (AI) system that can see through walls using wireless signals.
Researchers at The Massachusetts Institute of Technology's (MIT) Computer Science and Artificial Intelligence Laboratory showcased the RF-Pose system, which through the use of deep learning neural networks could be trained to spot human motion using RF signals reflected from people's bodies.
Trained through the use of both these signals and data correlated with on-camera movements, the AI could use just wireless signals - no radar or infrared - to create a heat map of movements and get an estimate of someone's movements and then mock them up as stick figures.
The tech works in poor lighting conditions and can also detect and represent multiple people; as the video below shows, it's pretty cool if a tad creepy (welcome to the world of AI tech).
For their next trick, the boffins plan to have the system create 3D skeletons to capture even smaller movements.
You might ask why this tech is needed outside of military applications or for the creation of robotic voyeurs, particularly as it could raise a few questions over privacy and the potential for spying.
But being able to see through walls and other objects without using X-rays or other radiation could be useful in search and rescue operations after natural disasters. like an earthquake or hurricane in an urban environment.
The tech also has potential in the medical world where it could be used to track the developing stages of movement related diseases such as Parkinson's, without requiring patients to wear obtrusive tech or have cameras dotted around their homes.
As the same MIT's clever folks see the tech as having the potential to keep an eye on elderly people who want to stay in their homes and live alone, but may need someone or indeed something to check on them in case they take a tumble or get in trouble.
But we probably won't see it until next year
Why stick a finger in a dyke when you can ram the entire boy in the hole, eh?
Reminds us that we're supposed to be able to trust them
'Exclusive' model starts shipping on 29 June