Researchers at MIT have come up with a system that tracks movement of individuals and medical disorders through radio waves.
Who says you need eyes to process vision or hard surfaces like walls to block sight? The researchers at the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) have reinvented movement tracking across walls, abandoning all mesh network or exotic radar tools along the way.
The pioneers have developed RF-Pose, a system which capitalises on neural network and radio signals to bounce off people, which in turn generates wireframe models in real time. In other words, the camera uses signals to sketch what lies on the other side of the wall without actually having a clear line of sight.
The neural networks have been cropping up in recent research endeavours and primarily work through training. The exhaustive set of data is labelled using humans, so that the radio images originating from the RF-Pose are accurately labelled for the computer to process the activity. The radio waves generated construct stick figures of people in a particular place, and can register multiple movements and micromovements at once. So, it not only identifies you when you’re walking or sitting, a mere action of writing will also be zeroed in on.
Considering the camera lacks the vision to process images, the system was primarily built for medical assistance–to detect the early growth of muscle and nerve disorders like Parkinson’s. However, the tacit hindrance of spying that comes with the radio waves bouncing off people poses vital privacy concerns. A refined version of the system that allows one to “opt in” to be traced through the range could possibly be configured down the line.
You can check out the intriguing workings of the neural network and radio waves through the system here: