Abstract: Various embodiments are disclosed for automated gait evaluation for retraining of running form using machine learning and digital video data. At least one machine learning routine is executed using video data to generate positional data of anatomical landmarks of a human or bipedal non-human subject in a video. Gait metrics and/or characteristics are determined for multiple stages of a gait cycle based on the positional data. An optimal gait cycle for a given body part or gait metric of the subject is determined and may be adjusted based on at least one covariable over time. A suggested change to movement patterns of the subject at specific timepoints or stages of the gait cycle that minimize a difference between the gait cycle of the subject and an optimal gait cycle are shown on a display device.

Abstract: Various embodiments are disclosed for automated gait evaluation for retraining of running form using machine learning and digital video data. At least one machine learning routine is executed using video data to generate positional data of anatomical landmarks of a human or bipedal non-human subject in a video. Gait metrics and/or characteristics are determined for multiple stages of a gait cycle based on the positional data. An optimal gait cycle for a given body part or gait metric of the subject is determined and may be adjusted based on at least one covariable over time. A suggested change to movement patterns of the subject at specific timepoints or stages of the gait cycle that minimize a difference between the gait cycle of the subject and an optimal gait cycle are shown on a display device.