Fisure 1. This figure illustrates the positions of the two accelerometers attached to the akin (highlighted with orange lines) and ine chest-mounted camera used for the MAR70+ dataset. (a) The buck acoelerometer was positioned centrally at the third lumbar vertebrae. The s-axis of the coondinate syatem was pointing forward. (b) The thigh accelerometer was positioned approximately 10cm above the upper border of the patella. The s-axis was pointing backward. The cameta was attached to the chest on the outside of the dothing, pointing downwards. Accelerometer data can be used to create a supervised machine learning model to predict human activities such as walking. sitting. standing and reclining based on accelerometers fixed to the body of a nursing home patients and a gopro camera. How can the raw data from the accelerometers be accurately labelled by researchers