Gait phase recognition using machine learning algorithm with IMU sensors Gait phase recognition is of great importance to develop accurate timing feedback for exoskeleton control. A powered exoskeleton can provide proper assistance to the subject if the current gait phase can be identified accurately. Machine learning classifiers were built in python to distinguish 4 gait phases in a gait cycle.