Fig. 1

Different representations of the IMU data for two example exercises. Example 1 was rated as a 1, and Example 2 was rated as a 3. The top left shows an example of a time-series representation of the unprocessed IMU data, the bottom left shows an example of an image representation of the IMU data, and the right shows example feature vectors hand-engineered from the IMU data. Example 1 and Example 2 were both taken from the training set. The numbers reported in the feature vector representation represent the statistical descriptors (e.g., root-mean-square of trunk sway in all directions, the path length of the trunk sway trajectory,and the elliptical fit area of the trunk sway ) for these examples used as input to the random forest model