Being able to classify movement signatures correctly is now pretty well achieved by most wearable or phone providers. Most of the MEMS sensors offer ready-to-use functions to detect walking, running, tap, double-tap or free-fall patterns but what if you want to detect specific movements for your niche market or for your custom application?
This is a companion discussion topic for the original entry at https://www.edgeimpulse.com/blog/classifying-movements-in-wearables-using-3-axis-accelerometers-and-machine-learning