The human mind is the most complicated object in the known universe.
Understanding how it works is the subject of intense research. FOCI is based on two decades of some of the most interesting research into cognition - neuro-respiration.
Researchers have established that our breathing pattern is closely correlated with our subconscious, cognitive states. When we consider how our breathing changes when we are stressed or relaxed, this seems intuitive.
If you are focused, or distracted, this will be reflected in the tiny movements in your breathing that the motion sensor detects. When a research scientist finds a neuro-respiratory pattern, this requires analyzing a huge data set, and hours of processing. We applied a powerful and agile form of machine learning to do this naturally, in real time - a 10g device, for everyday use.
In this demo, our learning engine understands that a triangle sitting on top a square is likely a house with just a few example - only a tiny fraction compared to conventional deep learning. Applied to FOCI, it reads waveforms from the motion sensor, and learns to differentiate real breathing from noise, see the pattern and match it to different cognitive states, most importantly adapting to you.
1. Respiratory Changes in Response to Cognitive Load: A Systematic Review. Neural Plasticity, 2016
2. Respiration-based emotion recognition with deep learning. Computers in Industry, 2017