It is called RF-Pose: this is the name given by MIT’s Computer Science and Artificial Intelligence Laboratory to a new technology that can see people through walls.
Using only artificial intelligence and radio waves, RF-Pose can reconstruct, in stylised form for the time being, whether a person on the other side of the wall is standing, sitting or walking. Based on artificial intelligence, the system can analyse the radio waves bouncing off people’s bodies. Since radio signals are ubiquitous, they were easier to use than other sensing technologies. How does RF-Pose work? The researchers trained the neural network by showing the video of a person walking next to the Radio signals he made as he moved; they then overlaid stick figures on the movement and trained the network to do the same automatically.
The system was trained both on examples of movement taken from video cameras and via signals reflected on people’s bodies, with the goal of making the system understand the association between a certain reflection and a given position.
According to the researchers, RF-Pose may be used especially for healthcare, to monitor the progression of diseases affecting movement like Parkinson’s and Multiple Sclerosis.