Machine Learning for Analyzing Human Movements in Sports

The modern-day athlete uses technology to receive intelligent feedback for improving technique and performance in competitive sports. Ideally, such analysis should take into account both the forces generated by the athlete (kinetics) and the resulting motions (kinematics). Initial Force AS has developed a software platform, called Motion Catalyst, which synchronizes information about human movements through high-speed video with three-dimensional force data obtained from a combined pressure and force plate.  

The proposed project will explore how to use machine learning on data captured through this platform to enrich the Motion Catalyst with intelligent feedback that could support coaches in guiding athletes towards improved performance. As part of this, the student(s) are expected to investigate existing AI and machine learning methods for assessing sports performance from kinetic and/or kinematic data. Furthermore, a desired outcome of the project would be the development of proof-of-concept machine learning models for analyzing the movements of professional athletes in data provided by Initial Force AS.

• Introduction to Motion Catalyst
• Data synchronization with Motion Catalyst
• More information: Motion Catalyst, Swing Catalyst
• More videos: SwingCatalyst@YouTube