These machine learning models are used to analyse walking gait from kinematic motion-capture data.

lr_decision_trees uses logistic regression and decision trees to predict when stance phase begins and ends. AutoDetect uses a convolutional neural network to predict when stance phase begins and ends.

AutoEncoder uses a convolutional neural network to fill gaps in statically or dynamically corrupted data.

AutoLabel uses two convolutional neural networks in parallel to automatically predict which data points correspond to which kinematic markers.